CO2 Conversion and Utilization 9780841237476, 9780841219076, 0-8412-3747-6

Citation preview

CO Conversion and Utilization

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.fw001

2

American Chemical Society Library 1155 16th St., N.W. In CO2 Conversion and Utilization; Song, C., et al.; Washington, D.C.20036

ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.fw001

A C S SYMPOSIUM SERIES

809

CO Conversion and Utilization Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.fw001

2

Chunshan Song, Editor Pennsylvania State University

Anne F. Gaffney, Editor Rohm and Haas Company

Kaoru Fujimoto, Editor The University of Kitakyushu (Formerly University of Tokyo)

American Chemical Society, Washington, D C

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Library of Congress Cataloging-in-Publication Data COO conversion and utilization / Chunshan Song, Anne F. Gaffney, Kaoru Fujimoto, editors. p. cm.—(ACS symposium series ; 809) Includes bibliographical references and index. ISBN 0-8412-3747-6

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.fw001

1. Carbon dioxide—Congresses. 2. Carbon dioxide—Environmental aspects— Congresses. I. Song, Chunshan. II. Gaffney, Anne F., 1954- III. Fujimoto, Kaoru, 1941. IV. American Chemical Society. Division of Petroleum Chemistry. V . American Chemical Society. Meeting (219 : 2000 : San Francisco, Calif.) VI. Series. th

QD181.C1 C52 546' 6812—dc21

2001 2001053943

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984. Copyright © 2002 American Chemical Society Distributed by Oxford University Press

A l l Rights Reserved. Reprographic copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Act is allowed for internal use only, provided that a per-chapter fee of $20.50 plus $0.75 per page is paid to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, M A 01923, U S A . Republication or reproduction for sale of pages in this book is permitted only under license from A C S . Direct these and other permission requests to A C S Copyright Office, Publications Division, 1155 16th St., N.W., Washington, D C 20036. The citation of trade names and/or names of manufacturers in this publication is not to be construed as an endorsement or as approval by A C S of the commercial products or services referenced herein; nor should the mere reference herein to any drawing, specification, chemical process, or other data be regarded as a license or as a conveyance of any right or permission to the holder, reader, or any other person or corporation, to manufacture, reproduce, use, or sell any patented invention or copyrighted work that may in any way be related thereto. Registered names, trademarks,etc.,usedinthis publication, even without specificindicationthereof,arenottobeconsideredunprotected

by

law.

QD

181

.C1C52 2002 c. 1

PRINTED IN T H E U N I T E DSTATESOFAMERICA

CO c o n v e r s i o n and utilization 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.fw001

Foreword The A C S Symposium Series was first published in 1974 to provide a mechanism for publishing symposia quickly in book form. The purpose of the series is to publish timely, comprehensive books developed from A C S sponsored symposia based on current scientific research. Occasionally, books are developed from symposia sponsored by other organizations when the topic is of keen interest to the chemistry audience. Before agreeing to publish a book, the proposed table of contents is reviewed for appropriate and comprehensive coverage and for interest to the audience. Some papers may be excluded to better focus the book; others may be added to provide comprehensiveness. When appropriate, overview or introductory chapters are added. Drafts of chapters are peerreviewed prior to final acceptance or rejection, and manuscripts are prepared in camera-ready format. As a rule, only original research papers and original review papers are included in the volumes. Verbatim reproductions of previously published papers are not accepted. A C S Books Department

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Preface Conversion and utilization of carbon dioxide, C 0 , is an important area of research for effective resource utilization and for sustainable development. C 0 is also a greenhouse gas whose concentration in the atmosphere has increased considerably in the last century. Consequently, C 0 utilization should be considered as an integral part of carbon dioxide management, because chemical conversion can not only add value to C 0 disposal, but also have environmental benefits. To foster the research and development in this area, we organized a symposium on C 0 Conversion and Utilization as a part of the 219th American Chemical Society (ACS) National Meeting, March 26-31, 2000, in San Francisco, California. This symposium was sponsored by the A C S Division of Petroleum Chemistry, Inc. This book was developed based on this A C S symposium. This book contains 25 peer-reviewed chapters that cover various aspects of C 0 conversion and utilization. Most of the chapters were developed from A C S papers presented at the above-mentioned symposium, and several chapters were invited contributions from leaders in the related research areas. The technical contents can be grouped into the following topics: (1) general overview, (2) synthesis of organic chemicals, (3) C 0 reduction over heterogeneous catalysts, (4) synthesis gas production from C 0 reforming, (5) effects of pressure and reactor type on C 0 reforming, (6) photocatalytic and electrochemical reduction, and (7) use of supercritical C 0 fluid. The individual chapters are contributed by active research groups from university, national laboratories, and industrial research organizations worldwide. The invited review chapters by leading experts cover the state of the arts in a wide variety of specific research topics, including C 0 mitigation and fuel production by Meyer Steinberg; organic chemicals synthesis using C 0 as a building block by Michèle Aresta; multifunctional catalysts for effective conversion of C 0 to valuable compounds by Tomoyuki Inui; C 0 emission reductions at the source by new catalytic technology by Leo Manzer; photocatalytic reduction of C 0 with H 0 by Hiromi Yamashita and Masakazu Anpo; and use of supercritical-C0 as a medium in catalytic reaction processes by Bala Subramaniam and Daryle H . Busch. A n introductory overview, which discusses the basic problems, scope, challenges, and opportunities of C 0 conversion and utilization, is provided by Chunshan Song. 2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.pr001

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

xi In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.pr001

A l l the contributions including the invited reviews were sent to two or three referees, and only the manuscripts that passed through the peer review process were incorporated in this book. As a result, some of the contributed manuscripts were not accepted, and most manuscripts were accepted after some revisions. We thank all the authors of the A C S book chapters, especially those who contributed to both the A C S book and the A C S symposium. We are very grateful to the authors of invited reviews, which substantially enhanced the breadth and depth of the scientific contents of this book. We express our sincere appreciation to all the peer reviewers whose time and efforts in evaluating the manuscripts have enhanced the quality of most chapters in the book, to the A C S Division of Petroleum Chemistry, Inc. for sponsoring this symposium, to the Department of Energy and Geo-Environmental Engineering of the Pennsylvania State University for the general support of this book project with respect to numerous mailings and communications, and to A C S Books Department for the skillful assistance of Kelly Dennis and Stacy VanDerWall in the acquisition process and of Margaret Brown in the editing and production process.

Chunshan Song Department of Energy & Geo-Environmental Engineering The Pennsylvania State University 206 Hosler Building University Park, P A 16802-5000

Anne M . Gaffney Rohm and Haas Company 727 Norristown Rd. Spring House, P A 19477-0904

Kaoru Fujimoto (Formerly University of Tokyo) Department of Chemical Process and Environments Faculty of International Engineering The University of Kitakyushu Wakamatsu-ku, Kitakyushu City, Fukuoka 808-0135, Japan

xii In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 1

C O Conversion and Utilization: An Overview 2

Chunshan Song

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

Clean Fuels and Catalysis Program, The Energy Institute, and Department of Energy & Geo-Environmental Engineering, The Pennsylvania State University, 209 Academic Projects Building, University Park, PA 16802

This article provides an introductory overview for defining the scope, potential and limitations of carbon dioxide (CO ) conversion and utilization. There are various sources of CO emissions, which are dominated by combustion of liquid, solid, and gaseous fuels. The amount of CO consumption for organic chemicals is relatively small compared to CO emitted from fossil fuel combustion. However, CO conversion and utilization should be an integral part of carbon management. Proper use of CO for chemical processing can add value to the CO disposal by making industrially useful carbon-based products. Studies on CO conversion into carbon-based chemicals and materials are important for sustainable development. CO conversion and utilization could also be positioned as a step for CO recycling and resource conservation. 2

2

2

2

2

2

2

2

2

2

Carbon dioxide, C 0 , is a colorless and odorless gas. The molecule is linear with a double bond between the carbon and oxygen atoms (0=C=0). C 0 occurs in nature and serves as source of carbon for photosynthesis of plants and crops. It is present in atmosphere with a volumetric concentration of 0.037 % (368 parts per million by volume) as of December 1999 (1). Combustion of most carbon-containing substances produces C O 2 . Energy utilization in modern societies today is'based on combustion of carbonaceous fuels, which are 2

2

2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

3 dominated by the three fossil fuels-coal, petroleum, and natural gas. Complete oxidation or combustion of any carbon-based organic matter produces C 0 , but until recently, C 0 gas was generally thought to be harmless. In fact, C 0 plays an important role in the Earth's carbon cycle, and is a necessary ingredient in the life cycle of animals and plants (2, 3). However, C 0 is also a greenhouse gas ( G H G ) , and the concentration of C 0 in the atmosphere has increased significantly (1,4, 5). There are increasing concerns for global warming and thus heightened interest worldwide for reducing the emissions of greenhouse gases, particularly C 0 (4, 5). This article is a general overview on C 0 emissions, C 0 conversion and utilization. The purposes of this overview are to define the scope, limitations, and potential of C 0 conversion and utilization, and to discuss the challenges and opportunities, because C 0 conversion and utilization should be an integral part of carbon management. Specific topical reviews are available on synthesis of organic chemicals by Aresta (6, 7), on chemical conversion of C 0 over heterogeneous catalysts by Inui et al. (8, 9), Arakawa (10) and Park et al. (11), on synthesis gas production from C 0 reforming of methane by Rostrup-Nielsen et al. (12, 13), Wang et al. (14) and Bradford and Vannice (15), on reactions in supercritical C 0 as a reaction medium by Leitner (16) and Subramaniam and Busch (17), on polymer synthesis and processing using supercritical C 0 by Cooper (18), on thermodynamics of chemical reactions by X u and Moulijn (19), and on various chemical reactions involving C 0 in the books by Paul and Pradier (4) and by Halmann and Steinberg (5). 2

2

2

2

2

2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

2

2

2

2

2

Sources of C 0 Emissions The sources of C 0 emissions include stationary, mobile, and natural sources, as listed in Table 1. The anthropogenic emissions include those from energy utilization in stationary and mobile sources, but exclude natural sources. Table 2 shows the worldwide C 0 emissions from the consumption of fossil fuels during 1980-1997 in million metric tons of carbon equivalent ( M M T C E ) (20, 21). Most countries in the world are consuming fossil fuels in stationary and/or mobile devices and thus contribute to C 0 emissions. Table 3 shows the U.S. emissions of C 0 from different sectors including residential, commercial, industrial and transportation sectors (22, 23). It is clear from Table 3 that every sector is contributing for C 0 emissions, even the residential homes are responsible because the electricity used in this sector is derived largely by fossil fuel-based electric power plants. During 1980-1997, C 0 emissions from U.S. industrial sector decreased somewhat (from 484.6 to 482.9 M M T C E ) , but those from transportation sector increased significantly, from 378.1 to 473.1 M M T C E . Within the industrial sector, there are two principal routes of C 0 formation, manufacturing of industrial products where C 0 is a byproduct from the process (such as the processes for production of cement, limestone, hydrogen, and ethylene oxide), and from energy supply (either as process heat or as electricity) by combustion of fossil fuels which produces C 0 . 2

2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

4

Table 1. Sources of Carbon Dioxide (CO2) Emissions Stationary Sources Mobile Sources Fossil Fuel-based Electric Cars, and Sports Utility Power Plants Vehicles

Natural Sources Humans

Animals Independent Power Trucks and Buses Producers Plant & Animal Manufacturing Plants in Aircrafts Industry Decay Land Commercial & Trains & Ships Emission/Leakage Residential Buildings Volcano Flares of Gas at Fields Construction Vehicles Earthquake Military & Government Military Vehicles & Facilities Devices a) Major concentrated C 0 sources include plants for manufacturing cement, limestone, hydrogen, ammonia, and soda ash as well as fermentation processes and chemical oxidation processes. a

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

5

Table 2. World C 0 Emissions from the Consumption of Fossil Fuels during 1980-1997 (in Million Metric Tons of Carbon Equivalent) 1997 1990 1980 World Regions 1725.8 1561.4 1484.2 North America 242.5 187.4 173.0 Central and South America 990.2 1011.2 1022.1 Western Europe 851.8 1297.9 Eastern Europe & Former U.S.S.R. 1111.4 264.0 200.8 137.2 Middle East 236.4 198.4 145.7 Africa 1921.1 1429.9 977.2 Far East and Oceania 6231.7 5886.9 5050.8 World Total Selected Countries 1488.5 1352.1 1290.8 United States N/A 1035.7 807.3 Former U.S.S.R. 421.8 N/A N/A Russia 821.8 620.4 393.0 China 296.7 273.6 260.4 Japan 234.4 N/A N/A Germany 78.4 N/A 82.6 Germany, East 188.8 N/A 207.7 Germany, West 167.4 156.9 167.7 United Kingdom 143.4 127.8 125.2 Canada 101.7 103.1 135.4 France 95.2 90.8 115.2 Poland 115.8 102.4 113.3 Italy N/A 79.7 Former Czechoslovakia 88.5 237.3 155.7 82.2 India 93.7 81.2 68.1 Mexico 98.9 80.5 64.2 South Africa 67.8 62.0 59.0 Spain 88.8 74.4 54.3 Australia 64.4 59.9 53.3 Netherlands 77.3 57.7 51.8 Brazil 31.3 49.5 46.8 Romania 36.8 34.2 37.4 Belgium 60.6 116.3 35.0 Korea, South 37.3 29.8 25.0 Venezuela 15.1 14.8 24.1 Sweden 67.1 40.0 23.2 Indonesia 44.6 35.0 18.0 Turkey 16.1 16.8 16.5 Austria 12.6 12.1 13.1 Switzerland Sources: DOE, EIA, 1998, 1999; N / A : not available, or not applicable.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

6 Table 4 shows the U.S. C 0 emissions from electricity-generating units that include both electric utilities and non-utilities (independent power producers) (22, 23). There are far more coal-fired units in the electric utilities, while natural gas-fired units are more popular among the independent producers. Overall, CO2 emissions from utilities and non-utilities contribute 523.4 and 134.4 M M T C E , respectively. Coal-fired units produced most CO2 (496.6 M M T C E ) , followed by gas-fired units (89.3 M M T C E ) and petroleum-based units (22.4 M M T C E ) . In addition to C 0 , there are several other greenhouse gases (GHGs). Table 5 shows the U.S. emissions of greenhouse gases during 1990-1997, based on a report published in 2000 by the Energy Information Administration of U.S. Department of Energy (24). Methane, nitrous oxides and fluorinated compounds have much stronger greenhouse gas effects than CO2. However, their amounts are much smaller. The emissions of all the GHGs are shown on the same basis using the same unit of million metric tons of carbon equivalent ( M M T C E ) based on global warming potential (24). The 1997 G H G emissions in the U.S. reached 1816 M M T C E , including contributions by C 0 (1505 M M T C E ) , C H (165 M M T C E ) , N O x (103 M M T C E ) and other greenhouse gases totaling 38 M M T C E . It is clear from these data that CO2 is by far the most dominant G H G . 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

4

Control of C O 2 Emissions Figure 1 shows the key issues related to greenhouse gas control and utilization. The basic issues of G H G control involve energy economics, policy regulations, environmental protection and global climate change. Control of CO2 emissions is among most important areas of G H G control. The anthropogenic emissions of G H G s are due mainly to the fossil energy utilization. The economic development in the U.S. has benefited greatly from relatively cheap energy sources. The idea for regulation of CO2 emissions by ratifying Kyoto Protocol has not been popular in the U.S., although emissions of pollutants (SOx, NOx, CO, H C , particulate matters, etc.) are being regulated strictly in the U.S. On the other hand, carbon tax has already been implemented for C 0 emissions in some countries in Western Europe. There are five broadly defined technical options for control of CO2 emissions: energy choices, energy efficiency, CO2 capture, CO2 sequestration, and CO2 utilization. These options are briefly discussed below. Energy choice is the basic consideration for selecting the primary energy input for new installations of energy systems or for switching between different forms of energy for existing energy systems. One can reduce the CO2 emission per million B T U by selecting the less C02-intensive form of energy, for example, natural gas versus coal. A n empirical rule of thumb is the atomic hydrogen-to-carbon ratio; the higher the H/C ratio, the lower the amount of CO2 per million B T U . The H/C ratio decreases in the order of natural gas (about 3-4), petroleum (about 1.8-2.0), and coal (about 0.8-1.2). Fuel decarbonization before combustion is an option that is being studied as one way to mitigate CO2 emissions to atmosphere (5), which is also related to the above discussion with respect to fuel H/C ratio changes. 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

7

Table 3. U.S. CQ2 Emissions from Different Sectors (Million Metric Tons of Carbon Equivalent) CO2 Emissions Sources 1990 1997 1980 CO2 C0 CO2 C0 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

from from from from

Residential Sector Commercial Sector Industrial Sector Transportation Sector

C 0 from End-Use Total 2

248.4 178.3 484.6 378.1

253.1 206.8 454.1 432.1

286.5 237.2 482.9 473.1

1289.4

1346.1

1479.6

C 0 from Electric Utilities* 476.9 418.4 Sources: D O E , EIA, 1998,1999 *Eleetrie Utility emissions are distributed across end-use sectors. 2

523.4

Table 4. U.S. C 0 Emission from Electricity-Generating Tons of Carbon Equivalent) CO2 Emissions Sources 1990 Coal-Fired Units at Electric Utilities 409.9 Petroleum-Fired Units at Electric Utilities 25.3 Gas-Fired Units at Electric Utilities 39.2 Other Units at Electric Utilities 1.2 Emissions at Electric Utilities, Sub-total 475.5

1995 434.3 13.0 44.5 0.8 492.7

1997 471.3 15.0 36.0 1.0 523.4

Coal-Fired Units at Nonutilities Petroleum-Fired Units at Nonutilities Gas-Fired Units at Nonutilities Other Units at Nonutilities Emissions at Nonutilities, Sub-total

17.8 4.3 39.2 37.4 98.7

24.6 7.3 57.6 45.9 135.5

25.3 7.4 53.2 48.4 134.4

C 0 from Coal-Fired Units, Total C 0 from Petroleum-Fired Units, Total CO2 from Gas-Fired Units, Total C 0 from Other Units, Total Total CO2 Emissions from Generators Sources: D O E , EIA, 1998, 1999

427.7 29.6 78.4 38.5 574.2

458.9 20.3 102.1 46.8 628.1

496.6 22.4 89.3 49.4 657.7

2

2

2

2

Units (Million Metric

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

8

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

Table 5. U.S. Emissions of Greenhouse Gases during 1990-1999 Based on Global Warming Potential (Million Metric Tons of Carbon Equivalent) 1999-P Gas 1997 1990 1995 1,527 Carbon Dioxide 1,505 1,351 1,435 Methane 172 165 182 179 104 103 Nitrous Oxide 99 106 24 38 HFCs, PFCs, and SF6 35 29 Total 1,833 1,816 1,655 1,748 a) Source: E I A Report, US D O E , EIA/DOE-0573(99), October 31, 2000

Figure 1. Key issues for control of greenhouse gas and related technical areas for CO2 control.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

9 A n alternative is to use more renewable energy. Common renewable energy forms include hydropower, solar energy, wind energy, and biomass. Biomass and animal wastes are opportunity fuels whose use is also encouraged where available. They are not completely C0 -neutral, because the collection and transportation of biomass consume energy and fuels. The issues o f major concerns on biomass include the regional and seasonal availability as well as energy density and energy demand-supply balance. Combination of biomass with fossil energy in certain ways including conversion is also being explored (25). Solar energy could also be used in selected endothermic chemical processes to reduce the need for using heat from combustion of fossil fuels which also produces C 0 . The issues of concerns for renewable energy in general are energy density, availability, energy efficiency, and capital cost. Development for improved energy efficiency is an important area that has a major impact on C 0 reduction. Existing energy utilization systems in the fossil fuel-based electricity generators in the U.S. have an average efficiency o f about 35% (26). In other words, about 65% of the useful energy input to the electric power units is wasted as conversion loss. The efficiencies for the automobiles are even lower, less than 20% (27). While significant improvements of energy efficiencies have been achieved in the past decades in power generation technologies and in transportation vehicles, there is a thermodynamic limit for the heat engines by Carnot cycle, which basically dictates the operation. Development and implementation of new energy utilization systems such as IGCC for coal-based power plants, G T C C for natural gas-based power plants, and fuel cell-based hybrid motors for transportation, could increase the energy efficiencies significantly, by 30% or more. The same principle can be applied to chemical industry, where more efficient process or more selective catalyst can make a process such as oxidation reaction more selective such that C 0 formation is minimized at the source, thus improving the process efficiency, conserving hydrocarbon resource, and reducing C 0 formation at the source (28). 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

2

2

C 0 Capture and Sequestration C 0 capture involves chemical or physical separation of C 0 from gas mixtures. Common methods include absorption using an agent such as monoethanol amine, physical adsorption using solid adsorbent, cryogenic separation at low temperatures, and membrane separation (see below). C 0 sequestration refers to long-term storage of C 0 in various reservoir locations with large capacity, such as geologic formations, ocean, aquifers, and forest (2931). The industrial separation of C 0 from flue gas o f power plants is currently carried out by using absorption process with monoethanol amine as the liquid absorbent (32, 33). It is usually carried out using gas mixtures that contain C 0 in relatively high concentrations. Large-scale separations in industry are currently based on chemical absorption of C 0 from concentrated gas mixtures such as flue gases from power plants, from hydrogen plants or 2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

10

from ammonia plants. The Fluor Daniel E C O N A M I N E F G CO2 recovery process that was developed by Dow Chemical is one of the widely used commercial processes (34). The process uses an amine solution, in which a proprietary additive is present, to remove C 0 economically from low-pressure, oxygen-containing streams such as flue gas. Some recent large-scale designs for CO2 recovery from flue gas are proposed for use in C02-enhanced oil recovery (35). It is known that for the conventional liquid scrubber, the sterically hindered amine requires considerably less energy for regeneration than conventional monoethanolamine (36). A team of researchers at Argonne National Laboratory has conducted a study to identify and evaluate the advantages and deficiencies of several technologies for capturing CO2 from the flue gas of utility boilers, including chemical solvent, cryogenic, membrane, physical absorption, and physical adsorption methods (37). Physical adsorption can be used for separation of gas mixtures. Activated carbons and carbon molecular sieves are readily available commercially, and many studies have been conducted on C 0 adsorption using such materials. A well-planned and carefully conducted study by Yang and coworkers at SUNY-Buffalo has demonstrated that P S A by using carbon molecular sieve could generate pure C 0 at low recovery, but generally the kinetic selectivity of C 0 / N is not high enough (38). Separation of CO2 from air is usually not economical, but there are special applications for such separation in spacecraft using solid sorbent. C 0 capture and sequestration (storage) are active areas of study worldwide, and recent reviews are available on the state of the art (29, 31, 33), on the options for C 0 removal from fossil-fueled power plants (39, 40), on U.S. D O E perspectives (41, 42), on the international perspectives (43), and on the membrane technologies for gas separation in general (44). A s indicated in most reviews, technologies for capturing CO2 are available but are still relatively energy-intensive and thus relatively expensive. Further research, particularly novel approaches, are necessary for lowering the cost of CO2 utilization. A n alternative approach has been proposed to use C 0 along with H 0 and O2 in flue gas for tri-reforming of natural gas without C 0 pre-separation (45).

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

2

2

2

2

2

2

2

2

C 0 Conversion and Utilization C 0 conversion refers to its transformation to chemically different forms that contain the carbon of CO2 or that makes use of active "oxygen atom" from C 0 . C 0 utilization includes the uses of CO2 in both physical processes such as extraction and chemical processes such as chemical synthesis. The objectives of research and development efforts on CO2 conversion and utilization can include one or more of the following specific goals depending on the applications: (1) to make use of C 0 for environmentally-benign physical or chemical processing based on the unique physical or chemical properties of C 0 ; (2) to produce useful chemicals and materials using C 0 as a reactant or feedstock; (3) to replace a hazardous substance or a less-effective substance in existing processes with C 0 as an alternate medium or solvent or co-reactant or 2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

11 a combination of them; (4) to use CO2 for recovering energy or for growing biomass while reducing its emissions to the atmosphere by sequestration; (5) to recycle C 0 as carbon source for chemicals and fuels; (6) to convert C 0 under geologic-formation conditions into "new fossil" energies. Table 6 shows the properties of carbon dioxide (2, 3, 46-48). CO2 may be used as a gas, liquid, and solid. A t atmospheric pressure, CO2 is about 1.5 times as heavy as air (2, 3). C 0 may be liquefied by compressing to 2 M P a and cooling to -18 °C, or by compressing to higher pressure of 5.78 M P a at 21 °C. If liquid C 0 is cooled further to -56.6 °C, the pressure drops to 0.518 MPa, solid C 0 is formed which co-exists with liquid and gaseous C 0 . This is known as the triple point. The solid C 0 can sublimate directly into gas (without going through the liquid phase) upon absorbing heat, and thus it is called dry ice. C 0 can not exist as liquid below the triple point. If the pressure on dry ice is reduced to atmospheric pressure, the temperature of dry ice drops to -78.5 °C. CO2 can not be liquefied above 31 °C, the critical temperature. It exists as supercritical fluid when the temperature and pressure are above 31 °C (Tc) and 7.38 M P a (Pc), respectively. Currently, CO2 is used as refrigerant for food preservation, beverage carbonation agent, inert medium (such as fire extinguisher), pressurizing agent, supercritical solvent, chemical reactant (urea, etc.), neutralizing agent, and as gas for greenhouses (2, 3, 46, 49). Solid C O (dry ice) has a greater refrigeration effect than water ice. Dry ice is also usually much colder than water ice, and the dry ice sublimates to a gas as it absorbs heat. It should be noted that the use of CO2 for refrigeration does not contribute to reduction of CO2 emissions. 2

2

2

2

2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

Thermodynamics of C 0 Conversion. Figure 2 illustrates the thermodynamics of CO2 conversion, where Gibbs free energy of CO2 and related substances are shown along with CO2. The data for plotting Figure 2 were taken mainly from two comprehensive chemical handbooks (47, 48). C 0 is a highly stable molecule. Consequently a substantial input of energy, effective reaction conditions, and often active catalysts, are necessary for chemical conversion of CO2. In other words, many reactions for CO2 conversion involve positive change in enthalpy, ΔΗ, and thus they are endothermic. However, it should be pointed out that the chemical reactions are driven by the difference in Gibbs free energy between the products and reactants at certain conditions, as shown by the equation below. 2

2

A G = ΔΗ - Τ AS There appears to be some perceptions by many people that C O 2 conversion would be so endothermic that its conversion would not be feasible. It is true that endothermic reactions consume energy. However, endothermic reactions can be feasible and indeed useful. There are many chemical manufacturing plants that are operated based on endothermic reactions in the chemical industry, and these include pyrolysis (thermal cracking) of

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

12

Table 6. Physical and Chemical Properties of Carbon Dioxide Property Value and Unit Sublimation point at 1 arm (101.3 kPa) - 78.5 °C Triple point at 5.1 atm (518 kPa) - 56.5 °C Critical temperature (T ) 31.04 °C Critical pressure (P ) 72.85 atm (7383 kPa) 0.468 g/cm or 468 g/L Critical density (p ) c

c

J

c

Gas density at 0 °C and 1 atm (101.3 kPa) Liquid density at 0 ° C and 1 atm (101.3 kPa) 25 °C and 1 atm C 0 (101.3 kPa) Solid density Specific volume at 1 atm and 21 °C Latent heat of vaporization At the triple point (- 78.5 °C) AtO°C Viscosity at 25 °C and 1 atm C 0 (101.3 kPa) Solubility in water at 0 ° C and 1 atm (101.3 kPa) 25 ° C a n d 1 atm (101.3 kPa) 2

2

1.976 g/L 928 g/L 0.712 vol/vol 1560 g/L 0.546 m /kg J

353.4 J/g 231.3 J/g 0.015 cP (mPas) 0.3346 g CO /100 g - H 0 or 1.713 m L C 0 / m L - H 0 at 0 °C 0.1449 g CO /100 g - H 0 or 0.759 m L C 0 / m L - H 0 at 25 °C - 393.5 kJ/mol 213.6 J/Knnol - 394.3 kJ/mol 2

2

2

2

2

2

Heat of formation at 25 °C Entropy of formation at 25 °C Gibbs free energy of formation at 25 °C

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

2

13

C6H6 (g) &

6

(+130)

C H (1) 6

C H (g) (+68) H (g) (0) N H (g) (-16) 2

4

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

6

(+124)

H (1) (+18) C Hi4 (1) (-4) 2

C-graphite (s) (0) Ni (s) (0)

6

3

CH4 (g) (-51)

CH3OCH3 (g)

ττρττη (-113)

H

C

H

O

m

(

iim 102

Φ (" >

CO (g) (-137)

C H 3 O H (1) (-166)

H 0 (g) (-237)

H 0 (1) (-228)

\ — C 0 (g) (-394)

C0

2

2

C6H5COOH (s) (-245)

2

HCOOH (1)

NH CONH (s) (-197) NiO (s) (-212)

2

(-361)

(1) (-386)

2

2

Ni(OH) (s) (-447) 2

C 0 = (1) 3

(-528)

HCO3- (I) (-586)

C 0 = (1) 2

4

M C O 3 (s)

(-613)

(-671)

CaC03 (s) (-1130) Gibbs Free Energy AG°f (KJ/moIe) (State, G: gas; L: liquid; S: solid) Figure 2. Gibbs free energy of formation for C 0 and related molecules. 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

14 hydrocarbons for manufacture of ethylene and propylene, dehydrogenation reaction for manufacture of petrochemicals such as styrene from ethylbenzene, and steam reforming of hydrocarbons for producing synthesis gas and hydrogen. A simple comparison between steam reforming of methane and CO2 reforming of methane can illustrate this point. Both reactions are endothermic that require over 200 kJ of energy input per mole of CH4, but CO2 reforming requires about 20% more energy input compared to steam reforming. The two reactions give synthesis gas products with different H / C O molar ratios; both are useful for certain applications. A careful analysis of the thermodynamic data in Figure 2 would reveal that it is more energy-demanding i f one were to use only C O 2 as a single reactant, but it becomes easier thermodynamically i f CO2 is used as a co-reactant with another substance that has higher Gibbs free energy, such as C H , carbon (graphite) and H . This trend can also be seen by the change in the reaction heat for reactions with C 0 as the single reactant [e.g., C 0 = C O + 1/2 0 ; ΔΗ° = + 293 kJ/mole C 0 ] and with C 0 as a co-reactant [e.g., C 0 +H = C O (g) + H 0 (g); ΔΗ° = + 51 kJ/mole C 0 ] .

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

4

2

2

2

2

2

2

2

2

2

2

Current Status and Potential Market of C 0 utilization. Table 7 shows the 1999 U . S . annual production of C 0 , C0 -based and C0 -related chemicals, as well as synthetic organic chemicals and materials (50), in which some 1999 data were derived by estimation based on 1994 data (51). Currently, the U.S. production of liquid and solid C 0 amounts to 5.36 million tonnes per year (Table 7), which is mainly consumed in the food and beverage industries (refrigerant, coolant, and carbonated drinks). In addition, there are about 10 million tones of C 0 consumed in making 4.95 million tones of urea fertilizers and 8.47 million tones of urea for chemicals synthesis (Table 7). The author did an analysis by calculation to set the potential upper limit of hypothetical C 0 demand for making organic chemicals and materials. Table 7 also shows the potential upper limit of CO2 demand estimated by the author for chemicals and materials. The methodology of estimation used here is based on that used by Steinberg (5) with modification by considering carbon equivalence in this work. Steinberg used the production data for the plastics in the U . S . in 1980 (5). The present estimation for the carbon-based synthetic materials is based on the U.S. production of plastics, fibers, and rubbers in the 1999 (50). If we assume that the potential upper limit of chemical market demand is that for making all the carbon-based synthetic materials plus several C0 -related major chemicals and materials using C 0 , then the upper limit of potential C 0 demand for synthetic organic chemicals and polymer materials could increase to 44.41 million metric tons of carbon equivalent (Table 7). This calculated numbers all correspond to a stoichiometric reaction system where all the carbon in C 0 is converted to the chemicals (urea, methanol, etc.) and synthetic polymer materials (plastics, fibers, rubbers). This is, of course, a hypothetical scenario, but it does provide an useful measure in terms of order of magnitude. 2

2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

15 Table 7. U.S. Production of Synthetic Plastics and Related Chemicals in 1999 and Estimated Potential Upper Limit of CO2 Demand for Chemicals and Materials Chemicals & Materials Production i n Common Metric Tons (Tonnes) Units 36,650,058 Synthetic Plastics: 80,727 millions of lb 4,639,426 Synthetic Fibers 10,219 millions lb Synthetic Rubbers 2,414 thousands of metric 2,414,000 tons 1. Polymers Subtotal 37,584,996 as C 43,703,484 as Comp (1999) = 13,579,604 Ammonia (reference 14,972 thousands of tons for urea) (1999) 4,945,871 Urea for fertilizer 5,453 thousands of tons (1999) ' 8,471,640 Urea for chemicals 18,660 millions o f l b (1999) ' Urea for chemicals 15.90 billions of lb (7,952 7,218,600 (1994) thousands of tons) 2. U r e a - e q u i v a l e n t 2,686,185 a s C 9,839,508 as C 0 C0 (1999)= 5,529,720 Methanol f o r 12.18 billions of lb chemicals (1994) 3. MeOH-Equivalent 2,428,590 as C 8,895,937 as C 0 C0 (1999r C0 - Liquid+Solid 11.80 billions of lb (5,899 5,357,200 (1994) ' thousands of tons) 1,711,143 a s C 4. Liquid+Solid C 0 6,267,924 as C 0 (1999) Total US P o t e n t i a l 44.4 M M T as C Ultimate U . S . C 0 162.8 M M T as C 0 demand for chemicals (1+2+3+4) = & materials Estimated for W o r l d 177.6 M M T as C Ultimate world C 0 651.3 M M T as C 0 demand for chemicals Potential & materials a) Source: A C S . Facts & Figures for the Chemical Industry. C & E N , June 26, 2000, 48-89; b) A significant fraction o f the urea is used for making thermosetting plastics that are included in synthetic plastics (2,691 millions of lb urea-based thermosetting resins were produced in 1999); c) Source: A C S . Facts & Figures for the Chemical Industry. C & E N , June 24, 1996, 38-79; d) Liquid and solid C 0 only; e) 1999 production of liquid+C02 was estimated (as 1.17 times 1994 production); f) 1999 production of methanol was estimated (as 1.17 times 1994 production); g) M M T : Million metric ton.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

a

a

a

b

a

b

b

c

2

2

c

2

2

2

A

c

d

2

e

2

ë

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

16 The upper limit of potential market demand worldwide for C 0 in making organic chemicals and materials is estimated to be 178 million metric tons of carbon equivalent. The detailed statistical data of chemicals production in 1999 are not available for all the countries in the world, but the author did a simplified estimation based on the U.S. chemical production. U . S . chemical industry is the largest in the world, and the plants in the U.S. provide 24% of the world's total chemical production (52). Consequently, the upper limit of potential market demand for CO2 in the world is estimated by the author to be 4 times that for the U.S. Table 8 shows the U.S. production of liquid fuels in 1999 (26). The amount of liquid fuels consumed in the U.S. has reached 13.60 million barrel per day, in which 12.75 million barrels per day liquid fuels were used for transportation fuels. B y simple calculation of carbon equivalence, it becomes apparent that the liquid fuels correspond to a very large amount of carbon that is on the same order of magnitude with the C 0 emitted from fossil fuel-based electric power plants in the U.S., as can be seen from Table 8. The total production of liquid fuels in the world is estimated to be about 4 times of the U.S. annual production, based on the fact that U.S. petroleum consumption is about 25% of the world total consumption (26). Table 9 shows the order of magnitude estimates by the author for C 0 utilization, based on the data shown in Tables 7 and 8. For comparison, Table 10 shows the order of magnitude estimates published in a recent review by Herzog for the worldwide capacity of various options for C 0 sequestration (31). The worldwide capacity for utilization of C Ô for making chemicals and materials is less than 1 Giga tonnes of carbon per year, which is relatively small compared to the several other options such as ocean sequestration. However, it should be noted that the amount of CQz that can be utilized for making chemicals and materials can be potentially large enough for expanding commercial-scale applications worldwide.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

2

2

2

Challenges and Strategies for C 0 Conversion and Utilization Figure 3 is an outline of possible chemical processes that may be used for C 0 conversion and utilization. There are non-catalytic chemical processes, processes using heterogeneous or homogeneous catalysis, photo-chemical and photo-catalytic reduction, bio-chemical and enzymatic conversion, electrochemical and electro-catalytic conversion, as well as solar-thermal/catalytic processes. Most of the processes are subjects of research in the laboratory, and few processes have reached large-scale production. A number of reviews have been published concerning various chemical reactions and processes for chemical conversion of C 0 (4-19, 43, 53, 54). There exist some chemical processes for C 0 conversion in chemical industry, for which synthesis of urea from ammonia and C 0 [ C 0 + 2 N H = H N - C O - N H + H 0 ] and the production of salicylic acid from phenol and C 0 [ C H - O H + C 0 = C H ( O H ) C O O H ] are representative examples. Urea is used for making various polymer materials and also for producing fertilizers. A s 2

2

2

2

2

2

2

6

5

2

2

2

3

2

6

5

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

17

Table 8. U.S. Production of Liquid Fuels in 1999

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

U.S. Fuels

1999 Daily Production

Million barrels per day Gasoline Distillate Fuels (Diesel, etc) Jet Fuel

8.38 3.55

1.67 Total = 13.60 [12.75 in transportation!

1999 Annual Productio η Million barrels per Year 3058.7 1295.8

Total Annual Production

609.6

C-Equivalent of Annual Prod c

Million Metric Tons (Tonnes)

M i l l i o n Metric Tons (Tonnes)

354.8 171.0

301.6 145.4

77.4 U.S. total =

65.8 512.8

W o r l d liquid fuel = 1997 US Electric Utilities Annual CO2 Emissions

2051.2 523.4 as C

a) Source: EIA, A E R , US DOE, 2000. b) 1 barrel of oil (US) = 42 US gallons = 0.15899 cubic meters (m ) = 158.99 liters. Assume average density values of the fuels at ambient temperatures as follows: 0.73 g/mL for gasoline; 0.80 g/mL for jet fuel; 0.83 g/mL for distillate fuels (diesel fuel and heating oils). c) Assume the average carbon content in the liquid fuels is 85 wt%. d) Estimated based on the fact that U.S. petroleum consumption is about 25% of the world's total consumption. 3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

18

Table 9. Order of Magnitude Estimates for the Worldwide Capacity of CO2 Utilization Option of C 0 Utilization Worldwide Capacity (Order of Magnitude in Giga Ton Carbon) Non-chemical Utilization 0.01-0.1 GtC per year Chemicals & Materials 0 . 1 - 1 GtC per year Synthetic Liquid Fuels 1 - 10 GtC per year a) Estimated by C. Song based on the U.S. and the world production of chemicals and materials as well as liquid fuels in 1999. 2

a

a

Table 1 0 . Order of Magnitude Estimates for the Worldwide Capacity of Various Sinks Sequestration Option Worldwide Capacity (Order of Magnitude in Giga Ton Carbon) Ocean 1000s GtC Deep saline Formations 100s-1000s GtC Depleted O i l and Gas 100s GtC Reservoirs Coal Seams 10s-100s GtC Terrestrial 10s GtC Utilization ( C h e m i c a l < 1 GtC per year Conversion) a) Source. H . J. Herzog. Using Carbon Capture and Sequestration technologies to Address Climate Change Concerns. A m . Chem. Soc. Div. Fuel Chem. Prepr., 2001,46(1), 53-55. a

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

19

Chemical/Catalytic

1

Photochemical/Catalytic

* Catalytic-Homogeneous 1

es 1 CO2 Conversion Processes f

Solar-thermal/Catalytic

»

Bio-chemical/Enzymatic

Electrochemical/Catalytic Figure 3. Chemical processes for CO2 conversion and utilization.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

20 an example of the usefulness of salicylic acid, acetyl salicylic acid is used for making Aspirin, a common medicine. Urea Synthesis C 0 + N H = H2N-CO-NH2 + H 0 2

3

2

Salicylic Acid Synthesis C H - O H + CO2 = C H ( O H ) C O O H 6

5

6

4

Barriers for C 0 Conversion and Utilization. There are significant barriers for promoting or enhancing utilization of C 0 in chemical processes. Such barriers and possible strategies to overcome them are briefly outlined below. 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

• 1) Costs of CQz capture, separation, purification, and transportation to user site. One of the strategies would be to identify and use concentrated CO2containing gas mixtures at or near the sites of chemical conversion and utilization. From the separation side, better separation methods and improved engineering of separation process can lower this barrier. However, i f it is beneficial and practically applicable, a new processing scheme for using C 0 in gas mixture without pre-separation of CO2 would be desired. A proposed new approach is to make industrially useful synthesis gas with desired H / C O ratios by using C 0 along with H 0 and 0 in flue gas for tri-reforming of natural gas without C 0 pre-separation (45). 2

2

2

2

2

2

• 2) Energy requirements of C 0 chemical conversion (plus source & cost of H if involved). The corresponding strategies are to use C 0 as a co-reactant along with one or more compounds that have higher Gibbs free energies, and to identify and use an effective catalyst that can give higher conversion at lower temperatures. C 0 reduction by hydrogénation has been studied and reported by many research groups. For such processes, H2 will need to be made from processes that do not co-produce C 0 . It should be noted that H is currently produced by reforming of hydrocarbons which is an energy-intensive process and accompanied by C 0 formation both from the conversion process and from the combustion of the fuels which is used to provide the process heat (54). 2

2

2

2

2

2

2

• 3) Market size limitations, and lack of investment-incentives for C0 -based chemicals. The estimated or perceived market sizes for C 0 conversion vary for different applications. If all carbon-based synthetic chemicals and materials can be made using C 0 , then the current annual production of such chemicals and materials could provide an estimated upper limit of market demands for C O 2 for chemicals and materials. 2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

21 • 4) Lack of socio-economical and political driving forces that facilitate enhanced C 0 utilization. There are several ways for an industrialized society to encourage more utilization of CO2. If an environment-conscious society wants to promote C 0 utilization, various incentives could be given to manufacturers that produce useful chemicals and materials using CO2, or make use of C 0 for environmentally-benign processing (such as replacement of phosgene). If the consumers and buyers in the industrialized societies recognize the need to such a level as to support C 0 utilization, price differential could be instituted for C 0 utilization which may be characterized as "carbon pricing for greenness". 2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

Strategy for Research on C 0 Utilization. The strategic directions for research and development on C 0 utilization may include but are not limited to the following considerations along with some specific examples. Figure 4 illustrates the utilization of C 0 with and without chemical conversion processing. 2

2

2

1) To produce useful chemicals and materials using C 0 as a coreactant or co-feed. Many processes could be designed for using C 0 , but the chemicals and materials that have large market demands are relatively limited. A n existing industrial process is urea synthesis, which has already found industrial application as fertilizer and as monomer for thermosetting polymers. Expanding applications of urea-based polymers can create more demand on C 0 and N H for urea synthesis. For the development of new processes, one of preferred approaches is to explore alternate processes for using C 0 as a co-reactant in making chemicals that have relatively large market or potentially large demand in the near future. Production of synthesis gas would fit in this category. Synthesis gas with desired H / C O ratios can be used for synthesizing chemicals such as methanol and acetic acid by oxo synthesis, olefmic chemicals and ultra-clean hydrocarbon fuels by Fischer-Tropsch synthesis, and also for electricity generation by using either the G T C C (gas turbine combined cycle) or hightemperature fuel cells such as molten carbonate fuel cells or solid-oxide fuel cells. Using CO2 and methane by dry reforming alone or in combination with steam reforming, and the synthesis of methanol and synthesis of hydrocarbon chemicals using C02-rich synthesis gas would fit in this category. Because of the problem of carbon formation and the limited applications of the synthesis gas with low H / C O ratios (around 1) from CO2 reforming, the combined reforming that involves both C 0 reforming and steam reforming may be preferred for large-scale applications (13, 45, 55). 2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

3

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

22

&

Chemicals Fertilizers

&

Solvent Medium

Fuels & Enhanced O&G Recovery

Figure 4. Utilization of C 0 with and without chemical conversion processing. 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

23 2) To make use of C 0 based on the unique physical or chemical properties of C 0 such as supercritical extraction using SC-CQ2 as a solvent. Environment-friendly and energy-efficient processes can be designed by using C 0 for separation and chemical reaction and materials synthesis. For example, supercritical CO2 can be used as either a solvent for separation or as a medium for chemical reaction, or as both a solvent and a reactant. Use of supercritical C 0 ( S C - C 0 ) allows contaminant-free supercritical extraction of various substances ranging from beverage materials (such as coffee bean), foods (such as potato chips), and organic and inorganic functional materials, to herbs and pharmaceuticals and soils. Some chemical reactions can benefit from using C 0 as a mild oxidant, or as a selective source of "oxygen" atoms. For example, the use of C 0 has been found to be beneficial for selective dehydrogenation of ethylbenzene to form styrene, and for dehydrogenation of lower alkanes such as ethane, propane and butane to form ethylene, propylene, and butene, respectively. Some recent studies have been reviewed by Park et al. on heterogeneous catalytic conversion using C 0 as an oxidant (11). 2

2

2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

2

• 3) To replace a hazardous or less-effective substance in existing processes with C 0 as an alternate medium or solvent or co-reactant or a combination of them. Some of the existing industrial chemical processes use hazardous substances as reactant or reaction medium which could be replaced by C 0 or supercritical C 0 . One area in this category is replacement of phosgene by C 0 , which has been discussed by Aresta for various chemical processes (56). Shown below is a comparison of different chemical processes for making dimethyl carbonate, an industrially useful chemical. U.S. phosgene use was estimated to be about 1.2 million tonnes per year (56, 57). In terms of environmental benefits, the new C0 -based route is superior to the existing industrial processes that are based on either phosgene or CO, as both chemicals are toxic. 2

2

2

2

2

Conventional Route (By SNPE, 1970s): C O C l + 2 CH3OH = CH3OCOOCH3 + 2 HC1 C O + C l = C O C l (Phosgene) 2

2

2

EniChem D M C Process (By EniChem - 12000 tons/Yr) C O + 1/2 0 + 2 CH3OH = C H O C O O C H + H 0 2

3

3

2

Ube D M C Process (By Ube Chemical - 3000 tons/Yr) C O + 2 R O N O = ROC(0)OR + 2 N O New C0 -Based Route C 0 + 2 CH3OH = CH3OCOOCH3 + H 0 2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

24 More recent studies by Dibenedetto and Aresta (58) have shown that direct oxidative carboxylation can be achieved using an olefin and CO2 under an oxidative condition, which can produce the cyclic carbonate chemicals that are currently synthesized in industry by using phosgene. 4) To use CO2 for energy recovery while reducing its emissions to the atmosphere by sequestration. C 0 in flue gas from power plants may be used for enhanced recovery of oil and natural gas, and for enhanced coal bed methane recovery. For enhanced oil recovery, the requirement for C O 2 purity is minimum, and thus the cost of gas processing (prior to use) is lower. Another approach in this direction is the so-called ocean fertilization which make use of C 0 in specific locations in the ocean and sun light to grow biomass (59).

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

5) To recycle C 0 as C-source for chemicals and fuels. Recycling of C 0 as carbon source for chemicals and fuels should be considered for applications where C 0 can be used that have desired environmental benefits. Although permanent storage seems preferred for carbon sequestration, C 0 recycling would also make sense i f such an option can indeed lead to less consumption of carbon-based fossil resources without producing more C 0 from the whole integrated system. The amount of liquid fuel consumed in the U . S . has reached 13.60 million barrel per day, which correspond to a very large amount of carbon that is estimated to be on the same order of magnitude with the C 0 emitted from fossil fuel-based electric power plants. The proposed approaches include hydrogénation of C 0 with H to methanol (60, 61), and conversion of C 0 in flue gas by tri-reforming to synthesis gas (CO + H ) with desired H / C O ratios for chemicals and fuels synthesis (45). 2

2

2

2

2

2

2

2

2

2

2

6) To convert C 0 under geologic-formation conditions into "new fossil" energies. Existing fossil fuels are believed to have originated from bio-chemical and geo-chemical transformations of organic substances that were initially present on the surface of the earth over the course of millions of years, which may be viewed as a detour in the carbon cycle. For example, numerous studies have shown that coal was formed from bio-chemical degradation and geochemical maturation of higher plant materials that were formed initially on the surface of the earth via photosynthesis from C 0 and H 0 . A new way of thinking has led to approaches of converting C 0 directly in geologic formations during its long-term storage. This is still a new topic for which only a few reports are available (62-64). 2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

25

Conclusions C 0 conversion and utilization is an important part of CO2 management strategy, although the amount of C 0 that can be converted to chemicals and materials is relatively small compared to the amount of anthropogenic CO2 emitted from fossil fuel combustion. C 0 is a carbon source and an unique substance, and C 0 utilization represents an important aspect of greenhouse gas control and sustainable development. Proper and enhanced uses of CO2 can have not only environmental benefits, but also add value to the C 0 disposal by making industrially useful carbon-based chemical products. There exist barriers and challenges for promoting C 0 utilization, which include but are not limited to the following: 1) Costs of C 0 capture, separation, purification, and transportation to user site; 2) Energy requirements of C 0 chemical conversion (plus source & cost of H2 i f involved); 3) Market size limitations, and lack of investment-incentives for CO2based chemicals; and 4) Lack of socio-economical and political driving forces that facilitate enhanced C 0 utilization. The following strategic considerations may be helpful for future directions: 1) To produce useful chemicals and materials using CO2 as a coreactant or co-feed; 2) To make use of C 0 based on the unique physical or chemical properties of C02 such as supercritical extraction using SC-CO2 as a solvent; 3) To replace a hazardous or less-effective substance in existing processes with C 0 as an alternate medium or solvent or co-reactant or a combination of them; 4) To use C 0 for energy recovery while reducing its emissions to the atmosphere by sequestration (storage); 5) To recycle C 0 as Csource for chemicals and fuels; and 6) To convert CO2 under geologicformation conditions into "new fossil" energies. 2

2

2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

2

2

2

2

2

2

2

Acknowledgments The author gratefully acknowledge the helpful discussions with Prof. Michèle Aresta of University of Bari, Italy, Prof. Tomoyuki Inui of A i r Water Inc. (formerly at Kyoto University), Japan, Dr. John Armor of A i r Products and Chemicals Inc., M r . Howard Herzog of Massachusetts Institute of Technology, Dr. David Beecy, Dr. Robert Kane, Dr. Curt White, Dr. Robert Warzinski and Dr. Donald Kraftsman of U.S. Department of Energy, Dr. Meyer Steinberg of Brookhaven National Laboratory, Prof. Alan Scaroni and Prof. Harold Schobert of Pennsylvania State University, M r . Wei Pan and Dr. Srinivas Srimat of Pennsylvania State University. The research on tri-reforming by the author received financial support by U C R Innovative Concepts Program of U . S . Department of Energy (DE-FG26-00NT40829).

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

26

References 1.

2.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

3.

4.

5.

6.

7.

Keeling, C. D.; Whorf, T. P. Atmospheric CO Records from Sites in the SIO A i r Sampling Network. In Trends: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U . S. Department of Energy, Oak Ridge, T N , U.S.A., August 2000. Lindsey, J. S. Carbon Dioxide. M c G r a w - H i l l Encyclopedia o f Chemistry, McGraw-Hill, Second Edition, New York, 1993, pp. 157159. Pierantozzi, R. Carbon Dioxide. Kirk-Othmer Encyclopedia o f Chemical Technology. Fourth Edition. V o l . 5, John Wiley and Sons,. New York, 1993, pp. 35-53. Paul, J . ; Pradier, C.-M. (Eds.). Carbon Dioxide Chemistry: Environmental Issues. Royal Society of Chemistry, Cambridge, U K , 1994, 405 pp. Halmann, M. M.; Steinberg, M. Greenhouse Gas Carbon Dioxide Mitigation: Science and Technology. Lewis Publishers, Boca Raton, F L , 1999, 568 pp. Aresta M. Perspectives of Carbon Dioxide Utilisation in the Synthesis of Chemicals. Coupling Chemistry with Biotechnology. Stud. Surf. Sci. Catal., 1998, 114, 65-76. Aresta, M. K e y Issues in Carbon Dioxide Utilisation as a Building Block for Molecular Organic Compounds in the Chemical Industry. In CO Conversion and Utilization. Song, C.; Eds.; Gaffney, A. M.; Fujimoto, K. American Chemical Society, Washington D C , 2001, in press. Inui, T.; Anpo, M.; Izui, K.; Yanagida, S.; Yamaguchi, T. (Eds.). Advances in Chemical Conversions for Mitigating Carbon Dioxide (Studies in Surface Science and Catalysis, V o l . 114), Elsevier, Amsterdam, 1998, 691 pp. Inui, T. Effective Conversion of CO to Valuable Compounds by Using Multi-functional Catalysts. In CO Conversion and Utilization. Song, C.; Gaffney, A. M.; Fujimoto, K.; Eds.; American Chemical Society, Washington D C , 2001, in press. Arakawa H . Research and Development on New Synthetic Routes for Basic Chemicals by Catalytic Hydrogenation of CO . Stud. Surf. Sci. Catal., 1998, 114, 19-30. Park, S.-E.; Y o o , J. S.; Chang, J.-S.; lee, K.Y.; Park, M. S. Heterogeneous Catalytic Activation of Carbon Dioxide as an Oxidant. A m . Chem. Soc. Div. Fuel Chem. Prepr., 2001, 46(1), 115-118. Rostrup-Nielsen, J.R. Aspects of CO -Reforming O f Methane. Stud. Surf. Sci. Catal., 1994, 81, 25-41. 2

2

8.

9.

2

2

10.

2

11.

12.

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

27 13.

14.

15. 16. 17.

Rostrup-Nielsen, J.R.; Hansen, Β. J.H.; Aparicio, L.M. Reforming of Hydrocarbons into Synthesis Gas on Supported Metal Catalysts. Sekiyu Gakkaishi, 1997, 40 (5), 366-377. Wang, S.B.; Lu G.Q.M.; Millar, G.J. Carbon dioxide reforming of methane to produce synthesis gas over metal-supported catalysts: State of the art. Energy & Fuels, 1996, 10 (4), 896-904. Bradford, M.C.J.; Vannice, M. A. CO reforming of CH . Catal. Rev.­ -Sci.Eng., 1999, 41 (1), 1-42. Leitner, W. Reactions in Supercritical Carbon Dioxide (scCO ). Top. Curr. Chem. 1999, 206, 107-132. Subramaniam, B.; Busch, D . H . Use of Dense-Phase Carbon Dioxide in Catalysis. In CO Conversion and Utilization. Song, C.; Eds.; Gaffney, A. M.; Fujimoto, K. American Chemical Society, Washington D C , 2001, in press. Cooper, A . I. Polymer Synthesis and Processing Using Supercritical Carbon Dioxide. J. Mater. Chem. 2000, 10 (2), 207-234. Xu, X.D.; Moulijn, J.A. Mitigation of CO by Chemical Conversion: Plausible Chemical Reactions and Promising Products. Energy & Fuels, 1996, 10 (2), 305-325. EIA/IEO. International Energy Outlook (IEO), Energy Information Administration, U.S. Department of Energy, DOE/EIA-0484(97), July 1998. EIA/IEO. International Energy Outlook (IEO), Energy Information Administration, U.S. Department of Energy, 1999. EIA/AER. Annual Energy Review (AER) 1997, Energy Information Administration, U.S. Department of Energy, Report No. D O E / E I A ­ -0384(97), July 1998. E I A / A E R . Annual Energy Review (AER) 1998, Energy Information Administration, U.S. Department of Energy, 1999. E I A / G H G . Emission of Greenhouse Gases in the United States 1999. Report No. EIA/DOE-0573(99), Energy Information Administration, US Department of Energy, October 31, 2000. Lee, K.-W. Clean Fuels from Biomass. A m . Chem. Soc. D i v . Fuel Chem. Prepr., 2001, 46(1), 61-64. EIA/AER. Annual Energy Review ( A E R ) 1999, Energy Information Administration, US Department of Energy, September 2000. D O E / E P A . Fuel Economy. Internet Resource Developed by U . S . Department of Energy and U.S. Environmental Protection Agency, http://www.fueleconomy.gov/feg/atv.shtml. Viewed May 20, 2001. Manzer, L. E. CO Emission Reductions: A n Opportunity for New Catalytic Technology. In CO Conversion and Utilization. Song, C.; Gaffney, A. M.; Fujimoto, K.; Eds.; American Chemical Society, Washington D C , 2001, in press. 2

4

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

2

18. 19.

20.

21. 22.

23. 24.

25. 26.

27.

28.

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

28 29.

30.

31.

32.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

33. 34.

35. 36.

37.

38.

D O E / O S - F E . Carbon Sequestration. State of the Science. Office of Science and Office of Fossil Energy, U . S . Department of Energy, February 1999. Herzog, H.J.; Drake, E.M. Carbon Dioxide Recovery and Disposal from Large Energy Systems. Ann. Rev. Energ. Env., 1996, 21, 145166. Herzog, H. J. Using Carbon Capture and Sequestration Technologies to Address Climate Change Concerns. A m . Chem. Soc. Div. Fuel Chem. Prepr., 2001, 46(1), 53-55. Chakma A . CO Capture Processes - Opportunities for improved energy efficiencies. Energy Conversion and Management, 1997, 38, S51-S56. Smith, I. M. CO Reduction - Prospects for Coal. I E A Report No. CCC/26, I E A Coal Research, London, U . K . , 1999, 84 pp. Sander, M.T.; Mariz, C.L. The Fluor Daniel E C O N A M I N E F G Proces - Past Experience and Present-Day Focus. Energy Conversion and Management, 1992, 33 (5-8), 341-348. Mariz C. L. Carbon Dioxide Recovery: Large Scale Design Trends. J. Canadian Petroleum Technology, 1998, 37 (7), 42-47. Mimura, T.; Shimojyo, S.; Suda, T.; Iijima, M.; Mitsuoka, S. Development of Energy-Saving Absorbents for the Recovery of Carbon-Dioxide from Boiler Flue-Gas. Kagaku Kogaku Ronbun, 1995, 21 (3), 478-485. Wolsky, A. M.; Daniels, E. J.; Jody, B. J. CO Capture from the Flue­ -Gas of Conventional Fossil-Fuel-Fired Power-Plants. Environ Prog. 1994, 13 (3), 214-219. Kikkinides, E. S.; Yang, R. T.; Cho, S. H. Concentration and Recovery of CO from Flue-Gas by Pressure Swing Adsorption. Ind. Eng. Chem. Res., 1993, 32 (11), 2714-2720. Gottlicher G, Pruschek R. Comparison of CO Removal Systems for Fossil-Fueled Power Plant Processes. Energy Conversion and Management, 1997, 38, S173-S178. Meisen Α.; Shuai, X.S. Research and Development Issues in CO Capture. Energy Conversion and Management, 1997, 38, S37-S42. Beecy, D . The Role of Carbon Sequestration in a National Carbon Management Strategy. A m . Chem. Soc. Div. Fuel Chem. Prepr., 2001, 46 (1), 38-44. Kane, R.; Klein, D. E. Opportunities for Advancements in Chemical Processes in Carbon Sequestration and Climate Change Mitigation. A m . Chem. Soc. Div. Fuel Chem. Prepr., 2001, 46(1), 275-277. Riemer, P. W. F.; Ormerod, W. G. International Perspectives and the Results of Carbon-Dioxide Capture, Disposal and Utilization Studies. Energy Conversion and Management, 1995, 36 (6-9), 813-818. Koros, W.J.; Mahajan, R. Pushing the Limits on Possibilities for Large Scale Gas Separation: Which Strategies?. J. Membrane Science, 2000, 175 (2), 181-196. 2

2

2

2

39.

40. 41.

42.

43.

44.

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

45.

(a) Song, C . Tri-reforming: A New Process for Reducing CO Emissions. Chemical Innovation (formerly Chemtech, A C S ) , 2001, 31 (1), 21-26; (b) Song, C. Chemicals, Fuels and Electricity Co-generated from Coal. A Proposed Concept for Utilization of CO from Power Plants. Proceedings of 16 Annual International Pittsburgh Coal Conference, Pittsburgh, P A , October 11-15, 1999, Paper No. 16-5. Ferguson, J. E . Carbon Dioxide. M a c M i l l a n Encyclopedia o f Chemistry. MacMillan References, New York, V o . 1, 1997, pp. 302-308. Lange's H D B K . Lange's Handbook o f Chemistry, 13 Edition, McGraw-Hill, New York, 1985, p. 9-4 and p. 9-70. C R C H D B K . C R C Handbook of Chemistry and Physics. 75 Edition, C R C Press, Boca Raton, F L , 1994, p.5-4. Ormerod, W.; Riemer, P.; Smith, A. Carbon Dioxide Utilisation. I E A Greenhouse Gas R & D Programme, ISBN 1 898373 17 5, IEA, London, U . K . , January 1995. A C S . Facts & Figures for the Chemical Industry. Chem. Eng. News, June 26, 2000, 48-89. A C S . Facts & Figures for the Chemical Industry. Chem. Eng. News, June 24, 1996, 38-79. OTP (Office of technology Policy, U.S. Department of Commerce.). Meeting the Challenge: U . S . Industry Faces the 21 Century. Chemtech, 1996, 26, 46-54. Audus H; Oonk H. A n Assessment Procedure for Chemical Utilisation Schemes Intended to Reduce CO Emissions to Atmosphere. Energy Conversion and Management, 1997, 38, S409-S414. Armor, J. N. Catalytic Fixation of CO , CO Purity, Energy, and the Environment. A m . Chem. Soc. Div. Petrol. Chem. Prepr., 2000, 45 (1), 141-142. Choudhary, V.R.; Mamman, A. S.; Uphade, B. S.; Babcock, R. E . CO Reforming and Simultaneous CO and Steam Reforming of Methane to Syngas over Co Ni O Supported on Macroporous Silica-Alumina Precoated with M g O . In CO Conversion and Utilization, Song, C.; Gaffney, A. M.; Fujimoto, K.; Eds.; American Chemical Society, Washington D C , 2001, in press. Aresta, M.; Quaranta, E. Carbon Dioxide: A Substitute for Phosgene. Chemtech, 1997, 27 (3), 32-40. Aresta, M.; Dibenedetto, Α.; Tommasi, I. Developing Innovative Synthetic Technologies of Industrial Relevance Based on Carbon Dioxide as Raw Material. Energy & Fuels, 2001, 15, 269-273. Dibenedetto, Α.; Aresta, M. Nb O as Catalyst in the Oxidative Carboxylation of Olefins. A m . Chem. Soc. Div. Fuel Chem. Prepr., 2001, 46(1), 119-121. 2

2

th

46.

47.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

48. 49.

50. 51. 52.

th

th

st

53.

2

54.

55.

2

2

2

2

x

1-x

2

56. 57.

58.

2

5

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

30

59.

60. 61.

62.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch001

63.

Markels, M.; Barber, R. T. Sequestration of Carbon Dioxide by Ocean Fertilization. A m . Chem. Soc. Div. Fuel Chem. Prepr., 2001, 46(1), 4548. Steinberg, M. CO Mitigation and Fuel Production. A m . Chem. Soc. Div. Petrol. Chem. Prepr., 2000, 45 (1), 74-76. Steinberg, M. CO Mitigation and Fuel Production. In CO Conversion and Utilization. Song, C.; Gaffney, A. M.; Fujimoto, K.; Eds.; American Chemical Society, Washington D C , 2001, in press. Beecy, D.; Ferrell, F. M.; Carey, J. K . Advanced Concepts for CO Conversion, Storage, and Reuse. A m . Chem. Soc. Div. Fuel Chem. Prepr., 2001, 46 (1), 99-100. Mahajan, D . ; Song, C.; Scaroni, A . W . Micro-reactor Study on Catalytic Reduction of CO into Liquid Fuels: Simulating Reactions under Geologic Formation Conditions. A m . Chem. Soc. Div. Petrol. Chem. Prepr., 2000, 45 (1), 113-117. Mahajan, D.; Song, C.; Scaroni, A . W. Catalytic Reduction of CO into Liquid Fuels: Simulating Reactions under Geological Formation Conditions. . In CO Conversion and Utilization. Song, C.; Gaffney, A . M.; Fujimoto, K . ; Eds. American Chemical Society, Washington D C , 2001, in press. 2

2

2

2

2

64.

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 2

CO Mitigation and Fuel Production 2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch002

M. Steinberg Department of Applied Science, Brookhaven National Laboratory, Building 475, Upton, NY 11973

CO mitigation technologies deals with how to utilize fossil fuels, coal, oil and gas with reduced CO emissions. Most development work to date has emphasized improving efficiency in generation and utilization of energy and in removal and recovery of CO from central power stations followed by disposal in underground wells or in the ocean. The latter suffers from economic penalties and potential adverse environmental effects. CO utilization for the chemicals industry is problematic because of the capacity mismatch between the gross CO emission and the relatively smaller chemical products market. However, utilization O f CO for conversion to alternative fuels for stationary and automotive power has potential of matching capacity between emissions and utilization. Although hydrogen-rich gaseous fuels such as methane and hydrogen can be used as alternative automotive liquid fuels which includes methanol and other higher oxygenates appear safer and fit in with the current liquid fuel infrastructure. The key for gas to liquid conversion utilizing CO is the production of hydrogen for conversion with CO . The Carnol process, catalytically reacts CO from coal fired plants with hydrogen from the thermal 2

2

2

2

2

2

2

2

2

The complete chapter is based on research from references 1 and 2.

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

32 decomposition of methane while sequestering the elemental carbon, to produce methanol and higher oxygenated fuels. CO emission reductions approaching 80% using fuel cell engines can be achieved compared to the conventional system of coal-fired power generating plants and gasoline-driven IC automotive engines.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch002

2

The Carnol Process The Carnol Process has been developed to convert C 0 from coal burning power plants to produce liquid automotive fuel with reduced C 0 emissions. The process is composed of three unit operations as follows: 2

2

1 . Carbon dioxide is extracted from the stack gases of coal fired power plants using monoethanolamine ( M E A ) solvent in an absorption-stripping operation. The technology for this operation is well known in the chemical industry for C 0 recovery and has recently been significantly improved for extracting C 0 from power plant stack gases. The power required to recover C 0 from an integrated coal fired power plant to recover 90% of the C 0 from the flue gas, has been reduced to about 10% of the capacity of the power plant. However, this energy requirement can be further reduced to less than 1% when the C 0 recovery operation is integrated with a methanol synthesis step described in item 3 below. 2

2

2

2

2

2. Hydrogen is produced with C 0 emission by the non-conventional method of thermally decomposing methane to carbon and hydrogen. 2

C H = C + 2H 4

2

The energy requirement in conducting this process per unit of hydrogen is less than that required by the conventional steam reforming process. A fluidized bed reactor has been used to thermally decompose methane and more recently attempts are being made to improve reactor design by utilizing molten metal or salt reactors. The carbon is separated and either stored or can be sold as a materials commodity, such as in strengthening rubber for tires. The temperatures required for this operation are 800°C or above and pressures of less than 10 atm.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

33 3. The third step in the process consists of reacting the hydrogen from step 2 with the C 0 from step 1 in a conventional gas phase catalytic methanol synthesis reactor. 2

C 0 + 3H = CH OH + H 0 2

2

3

2

This is an exothermic reaction so that the heat produced in this operation can be used to recover the stack gas C 0 from the absorption/stripping operation described in step 1, thus reducing the energy required to recover the C 0 from the power plan to less than 1% of the power plant capacity. This is an advantage compared to the energy cost in terms of derating the power plant when C 0 is disposed of by pumping liquid C 0 into the ocean in which case more than 20% of the power plant capacity is consumed. The gas phase methanol synthesis usually takes place at temperatures of 260°C and pressure of 50 atm using a copper catalyst. The synthesis can also be conducted in the liquid phase by using a slurry of zinc catalyst at a lower temperature of 120°C and 30 atm hydrogen pressure. 2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch002

2

2

2

In its simplest form the Carnol Process is a two step operation. When hydrogen is used to supply the energy for the thermal decomposition of methane, then the C 0 emission for methanol production is reduced to zero. 2

Methanol and Higher Oxygenates as Liquid Automotive Fuel The Carnol Process can be considered as a viable coal-fired C 0 mitigation technology because the resulting large production capacity of liquid methanol resulting from the large amount of C 0 emitted can be utilized in a large capacity automotive fuel market. Most processes which utilize C 0 produce chemical products which tend to swamp the market and thus cannot be used effectively. Methanol as an alternative automotive fuel has been used in internal combustion (IC) engines as a specialty racing car fuel for a long time. The E P A has shown that methanol can be used in IC engines with reduced C O and H C emissions and at efficiencies exceeding gasoline fuels by 30%. Methanol can also be used either directly or indirectly in fuel cells at several times higher conversion efficiency for automotive use. A great advantage of methanol is that, as a liquid, it fits in well with the infrastructure of storage and distribution compared to compressed natural gas and gaseous or liquid hydrogen, which are also being considered as alternative transportation fuels. Compared to gasoline, the C 0 emission from methanol in IC engines is 40% less and over 70% less for fuel cell vehicles. However, other higher oxygenate fuels such as dimethyl 2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

34 oxide which can be produced by catalytic Fisher-Tropsch reactions can be incorporated in the Carnol System. The thermodynamics ο f these reactions and C 0 emissions are shown in Table I. The entire Carnol System is shown in Figure 1. 2

It should also be pointed out that removal and ocean disposal of C 0 is only possible for large central power stations which for the U.S. accounts for only 30% of the total C 0 emission. For the dispersed industrial, domestic and transportation (automobiles) sectors, the Carnol Process provides the capability of C 0 reduction in these sectors by supplying liquid methanol fuel to these numerous small dispersed C 0 emitting sources. 2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch002

2

2

2

Economics of Camol Process At $ 18/bbl oil, 90% refining gasoline yield and $10/bbl for refining cost, gasoline costs $0.78/gal, and methanol being 30% more efficient than gasoline, competes with gasoline at $0.57/gal methanol. Currently, the market for methanol is depressed because of over supply due to removal of mandatory requirements for M T B E oxygenation of gasoline. In terms of the cost of reduction of C 0 from power plants, with S2/MSCF natural gas, and a $0.55/gal methanol income the C 0 reduction cost is zero. A t S3/MSCF natural gas and $0.45/gal income from methanol, the C 0 reduction or disposal cost is $47.70/ton C 0 , which is less than the maximum estimated for ocean disposal. More interesting, without any credit for C 0 disposal from the power plant, methanol at $0.55/gal can compete with gasoline at $0.76/gal (~$18/bbl oil) when natural gas is at $2/MSCF. Any sale of elemental carbon reduces the cost for reducing C 0 emissions. 2

2

2

2

2

2

C 0 Emission Evaluation of Entire Carnol System 2

The entire Carnol System is evaluated in Table II in terms of C 0 emissions and compared to the alternative methanol processes and to the base line case of conventional coal fired power plant and gasoline driven automotive IC engines and in fuel cell engines. A l l the cases are normalized to emissions from a 1 M M B T U of coal fired power plant, which produces C 0 for a Carnol methanol plant equivalent to 1.27 M M B T U for use in automotive engines. The assumptions made are listed at the bottom of Table II The conclusions drawn from Table II are as follows: 2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

3

3

2

2

2

6

34

2

2

2

2

2

13,340 59 23 -67 -51 46 2C0 +5H =C H 0+2H 0 Dimethyl Ether C * Fuel and water in liquid state * * C 0 emission is only for the conversion process. The emission from a conventional methane steam reforming plant for H production ranges from 177 to 205 Lbs C 0 / M M B T U of oxygenated or hydrocarbon fuel which is 8 times higher than for the thermal decomposition of methane for hydrogen production shown above.

2

16

65

2

20,000

2

25

6

-821

62

-149

15,800

62

226

2

24

14,580

61

16C0 +49H =C H +32H 0

-180

24

12,750

Diesel Ci -Cetane

-80

-130

24

64

74

-74

-82

20,600

2

2

60

-66

56

25

18

9

2

46

10,000

-396

8

4

7

2

22

- 60

2

2

3

5

-31

-57

32

Carbon Sequestered Lbs C/MMBTU

114

2

2

2

2

4C0 +12H =C H OH+7H 0

2

2

3C0 +9H =C H 0H+5H 0

2

2

2C0 +6H =C H OH+3H 0

2

Heat of Com bus. AHQ BTU/Lb

8C0 +25H =C H +16H 0

4

Reaction

1C0 +3H =CH OH+H 0

CO2 Emission** Lbs CO2/MMBTU

Gasoline Cg-Octane

c

Butanol

c

Propoanol

c

Ethanol

Ci

Methanol

Fuel

R

HT. of React. AH AHQ Kcal/Mol

2

HT. of Form AHF* Kcal/mol

2

Mol Wt of Fuel

2

Process: H Produced by Thermal Decomposition of Natural Gas (CH4 = C + 2 H ) at 80% Efficiency and Catalytic Reaction with C 0 to Produce the Fuel Indicated

2

Table I. Utilization of C 0 for Oxygenates and Hydrocarbon Fuels

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch002

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

2

1) 90% recovery of C 0 from coal fired plant. 3) Fuel cell is 2.5 times more efficient than conventional gasoline IC engine. 5) Only 53% emissions of coal plant C 0 is assigned to Carnol for fuel cells.

5

2

z

89

3

Fuel Cell

175

175

175

175

285

IC Automotive Power Plant

117

219

228

390

448

515

Total System Emission 2

2

77%

57%

56%

24%

13%

0%

C0 Emission Reduction

2) Methanol is 30% more efficient than gasoline in IC engine. 4) Only 25% reovery of C 0 from coal plant is necessary for supplying C 0 to conventional methanol plant.

17

ll

Case 4 Coal Fired Power Plant with C A R N O L Methanol and Fuel Cell Automotive Power

32

54

56

15

Fuel Process Plant

43

2

Case 3 Coal Fired Power Plant with Biomass for Methanol Plant

Case 2 Coal Fired Power Plant with C A R N O L Process Methanol Plant

4

1

161

Case I B Coal Fired Power Plant with C 0 Addition to Conventional Methanol Plant

21

215

Case 1A Coal Fired Power Plant with Conventional Steam Reformed Methanol Plant

2

215

Coal Fired Power Plan

Basline Case; Coal Fired Power Plant and Gasoline Driven IC Engine

System Unit

Fuel Process Plant and Automotive Power Plant

Basis: 1 M M B T U for coal fired 900 MW(e) power plant 1.27 M M B T U o f liquid fuel for IC engine - other fuel efficiencies proportions energy up and down C 0 2 Emission units in lbs C 0 / M M B T U (multiply by 0.43 for K G / G J )

2

Table II. C 0 Emission Comparison for Systems Consisting of Coal Fired Power Plant,

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch002

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Figure L Integrated System For CO2 Emission Reduction Carnol Process

METHANOL PLANT COFFREE

2

CARNOL CHEMICAL PLANT

Carbon

COAL-FIRED POWER PLANT

CO2 Flue Gas

ELECTRICITY CQ -FREE EMISSION

Electricity

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch002

AUTOMOTIVE FUEL USE REOUCEO CO; EMISSION

38 1. The use of conventional methanol reduces CO by 13% compared to the gasoline base case and is mainly due to the 30% improved efficiency of the use of methanol in IC engines. 2

2. B y addition of CO recovered from the coal fired power plant to the conventional methanol process, the CO from the power plant is reduced by about 25% (161 l b s / M M B T U compared to 215 lb CO /MMBTU) and the CO emissions for the entire system is reduced by 24%. It should be pointed out that part of the CO can also be obtained from the flue gas of the reformer furnace of the methanol plant. 2

2

2

2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch002

2

3. The Carnol Process reduces the coal fired power plant C 0 emission by 90% and the overall system emission is reduced by 56%. 2

4. Since the use of biomass is a C 0 neutral feedstock, there is no emission from the power plants because the biomass feedstock comes from an equivalent amount of C 0 in the atmosphere which has been generated from the coal fired power plant. Thus, the only net emission comes only from burning methanol in the automotive IC engine and thus, the C 0 emission for the entire system is reduced by 57%, only slightly more than the Carnol System. However, the main point is that, at present, the cost of supplying biomass feedstock is higher than that of natural gas feedstock. 2

2

2

5. Another future system involves the use of fuel cells in automotive vehicles. The efficiency of fuel cells is expected to be 2.5 times greater than gasoline driven engines. Applying the Carnol Process to produce methanol for fuel cell engines reduces the C 0 emission for the entire system by as much as 77%. Furthermore, because of the huge increase in efficiency, the capacity for driving fuel cell vehicles can be increased by 92% over that for conventional automobiles using the same 90% of the C 0 emissions from the coal burning power plant. 2

2

REFERENCES 1.

Meyer Steinberg "CO Mitigation and Fuel Production", B N L 65454, Brookhaven National Laboratory, Upton, NY (1997).

2.

Meyer Steinberg "Efficient Natural Gas Technologies: A Response to Global Warming", Chemtec 31-36 (January 1999).

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 3

CO Emission Reductions: An Opportunity for New Catalytic Technology Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch003

2

Leo E. Manzer DuPont Central Research and Development, Experimental Station, Wilmington, DE 19880-0262

Summary The theme of this symposium is related to reducing carbon dioxide emissions by conversion of the CO to other products. While this is a worthwhile object, the real goal where possible, should be to avoid the production of CO in the first place. It is the purpose of this paper to review processes that generate CO and provide some opportunities for research, which ultimately will reduce CO at the source to avoid end of pipe treatment. 2

2

2

2

Introduction Catalytic oxidations are among the least selective of all catalytic reactions and the source of much of the carbon dioxide from chemical processes. These processes are often operated at very high temperatures, with selectivity to desired oxygenated products of less than 90%. The major byproduct is usually carbon dioxide. Table 1 shows typical selectivity to the major product in these large-scale commercial operations. Most of these processes have production capacities of several hundred million lb./year, so the amount of C 0 generated is quite significant. For example, for a maleic anhydride plant operating at a capacity of 200-MM lb./year, with 60% selectivity, over 500-MM lb./year of 2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

39

40

carbon dioxide is produced. Therefore, there is a large incentive to improve yield to the hydrocarbon product in these processes. This can be done by improving the catalyst in these existing processes or by completely changing chemistry and engineering of existing processes. This paper will provide a brief overview of current trends and opportunities to reduce CO2.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch003

Table 1. Selectivity to Major Product for a few Commercial Catalytic Oxidation Reactions Oxidation Process Butane Oxidation Propylene Oxidation Propylene Ammoxidation Ethylene Oxidation

Major Product Maleic Anhydride Acrolein/Aery lie Acid Acrylonitrile Ethylene Oxide

Selectivity 60% 75% 80% 88%

Discussion

2.1 Anaerobic vs. Aerobic Oxidations It has been known for many years that in certain oxidation reactions the most selective catalysis occurs when the oxygen in the final product is derived directly from the lattice of the oxide catalyst. However, most large-scale commercial processes are carried out in the non-flammable region, which often requires a feed of less 99%. It was found that Z r 0 is an effective catalyst for the D M C synthesis from C H O H and C 0 . More details were described in our previous reports (12,13). 3

2

2

3

2

2

2

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch005

2

2

3

2

The effect of modified Z r 0 catalysts calcined at 673 Κ and 923 Κ on D M C formation are shown in Table II and III, respectively. The modification with some phosphates and sulfates promoted the D M C formation. H P 0 was found to be the most effective for the enhancement of the catalytic activity in this reaction at both calcination temperatures (14). D M E was formed on C e ( S 0 ) / Z r 0 and F e ( S 0 ) / Z r 0 calcined at 923 K . This suggests that these catalysts have stronger acidity. Figure 1 shows the dependence of D M C amount on reaction temperature over Z r 0 and H P 0 / Z r 0 (P/Zr = 0.025) calcined at 673 K . D M C formation on Z r 0 is controlled by the reaction rate under these reaction conditions. The dramatic effect of the modification with H P 0 was easily observed at all reaction temperatures. It seems that the amount of D M C on H P 0 / Z r 0 (P/Zr = 0.025) at 443 Κ reached the equilibrium level. The D M C amount on H P 0 / Z r 0 (P/Zr = 0.025) at 423 Κ became about four times larger than that on Z r 0 . The amount of D M C on H P 0 / Z r 0 (P/Zr = 0.025) at 403 Κ was comparable to that on Z r 0 at 443 K . While D M C formation was not observed on Z r 0 at 383 K , 0.10 mmol D M C was formed on H P 0 / Z r 0 (P/Zr = 0.025). It should be noted that the addition of H P 0 to Z r 0 promoted this reaction remarkably (14). Figure 2 shows the dependence of D M C amount and B E T surface area of H P 0 / Z r 0 calcined at 673 Κ on H P 0 loading. The amount of D M C and the surface area increased with the loading of H P 0 in the range of low P/Zr. The amount of D M C reached a maximum at P/Zr = 0.05, and then decreased at P/Zr > 0.05. Especially, H P 0 4 / Z r 0 (P/Zr = 0.3, 0.5) had high surface area, but exhibited very low activity for D M C formation (14). X R D patterns of fresh H P 0 / Z r 0 with various H P 0 loadings are shown in Figure 3. On Z r 0 , both metastable tetragonal phase and monoclinic phase were observed. Metastable tetragonal phase gradually decreased and monoclinic 2

3

2

4

3

2

2

2

3

4

4

3

4

2

2

2

3

4

3

3

4

4

2

2

3

4

2

2

2

3

3

3

4

2

3

4

2

4

3

3

4

2

4

2

3

4

2

3

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

76

Table I. Results of C H O H + C 0 reaction over heterogeneous catalysts 3

2

Amount of formation I mmol Catalyst at 673 K ) Z r 0 (calcined Z r 0 (commercial) Si0 A1 0 Ti0 H-ZSM-5 H-USY H-Mordenite 2

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch005

2

2

2

3

2

Η-β

MgO CaC0 SrC0

3

3

Y2O3

Hf0 Nb 0 2

2

5

M0O3

Sn0 ZnO Ga 0 Ge0 ln 0 Sb 0 Bi 0 La 0 PrrAi MgW0 CaW0 N i S 0 (calcined at 573 K ) 2

2

3

2

2

3

2

3

2

3

2

3

3

3

4

PMC 0.30 0.22 n. d. n. d. n. d. n. d. n. d. n.d. n. d. n. d. n. d. n. d. n.d. n.d. n. d. n. d. 0.14 n. d. n. d. n. d. n. d. n. d. n. d. n. d. n. d. n. d. n. d. n. d.

PME n. d. n. d. n.d. 0.11 0.16 72.1 11.3 11.6 26.9 n. d. n. d. n. d. n. d. n.d. n. d. n. d. n. d. n.d. n. d. n. d. n. d. n. d. n.d. n. d. n. d. n. d. n.d. 0.39

BET surface area 2

1

I 118 m g' 92 380 100 50 225 377 239 450 37 1 6 6 6 5 1 25 4 22 -

9 2 1 7 _

99.999%) and C 0 (>99.995%) was passed over the catalyst at 850°C. Partial pressure of CH4 or C 0 , denoted as P(CTL0 or P(C0 ), was 30.3 or 70.7 kPa, respectively, unless otherwise stated. The effluent after removal of H 0 was sampled at 5-min intervals, and C H , C H4, CO and H as products were analyzed with a high-speed micro GC. GC data were processed on the assumption that the C in CH4 was converted to C H , C H4, and CO, and the C in C 0 to CO. Since CO is formed from CH4 and C 0 , each contribution is separated by the previous method (1) in which almost all of side reactions involving CH4 and C 0 are taking into account. ŒU conversion and C selectivity can thus be calculated by the following equations (6) and C yield is defined as the product of both values. 2

2

2

2

2

2

6

2

2

6

2

2

2

9

2

2

9

2

CH4 conversion (%) = {2[C H ] + 2[C H4] + [CO from CH4D/UCH4] + 2[C H ] + 2[C H4] + [CO from CTLJ} χ 100 2

2

6

6

2

2

C selectivity (%) = {2[C H6] + 2[C H4]}/{2[C H6] + 2[C H4] + [CO from CH4]}xlOO 2

2

2

2

2

Catalyst Characterization Fresh and used catalysts were characterized by several methods, such as N adsorption at 77 K, X-ray diffraction (XRD) with Ni-filtered Cu K radiation, and X-ray photoelectron spectroscopy (XPS) with Mg K radiation. Binary a

a

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

89 catalysts after 2 h reaction of CH4 and C 0 at 850°C were also subjected to the C 0 temperature programmed desorption (TPD) measurements. In a TPD run, the catalyst after reaction under PiŒU) of 30.3 kPa and P(C0 ) of 70.7 kPa was first quenched to 100°C in a stream of feed gas with the same composition, then held for 30 min at this temperature after replacement with pure He, and finally heated at 2.5°C/min up to 950°C under flowing a mixture of He and C 0 with different PCCCb) in the range of 0 - 70.7 kPa. It took about 5 min for quenching the used catalyst to 100°C. The change in C 0 concentration during the TPD run was monitored with the micro GC. The reproducibility of the change at a desorption peak was within ±5% under the highest P(C0 ) of 70.7 kPa. 2

2

2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

2

2

Results and Discussion Catalyst Composition The effect of catalyst composition on the performance of binary Ca-Ce catalysts under P(CH0 of 30 kPa and P(C0 ) of 30 kPa is illustrated in Figure 3, where the composition is expressed as [Ca/(Ca + Ce) χ 100] in atomic percent (at%). CH4 conversion was 12% or 0.1% with Ce0 or CaO alone, respectively. The conversions over the binary catalysts were lower than the arithmetic mean of those over each component. C selectivity was nearly zero with Ce0 alone, but it steeply increased when a low content of CaO coexisted. C selectivity was higher over the binary catalysts than over CaO alone. As shown in Figure 3B, C yield was < 0.1% with Ce0 or CaO alone, whereas it was much higher in the coexistence of both components. In other words, there was a synergistic effect on C formation (4,6). The maximal yield reached 3.2%, which was about 30 times that observed with the single oxide. Table 1 summarizes the results when Ce0 power was impregnated with other nitrate solutions than Ca(N0 ) . Surface areas of the binary oxides prepared, determined by the BET method, were almost independent of the kind of the alkaline earth metal. CH4 conversion at 850°C increased in the sequence of Ba - Sr < Ca < Mg, whereas C yield did in the order of Mg < Ba - Sr < Ca (6). Thus, the Ca-Ce system showed the highest activity for C formation. When CaO and Ce0 were physically mixed at Ca/Ce ratio of 0.5, as shown in Table 1, surface area and CH4 conversion were almost the same between the resulting mixture and the corresponding binary catalyst. On the other hand, C yield was much higher with the latter. These observations indicate that catalyst preparation by the impregnation method is more effective for C formation. 2

2

2

2

2

2

2

2

2

3

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

90

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

91 On the basis of the results described above, the impregnation method using Ca(N0 ) solution was used to prepare a binary catalyst containing Cr or Mn element. Figure 4 shows the effect of catalyst composition on CH4 conversion at 850°C with the Ca-Cr (5) or Ca-Mn system. The conversion - composition profiles observed were quite similar as that (Figure 3) over the Ca-Ce catalyst; CH4 conversion with C r 0 or M n 0 alone decreased steeply by addition of a small amount of CaO and it leveled off with further increase in CaO. As is seen in Figure 5, C yield with C r 0 or M n 0 alone was low. On the other hand, Ca-Cr and Ca-Mn catalysts enhanced remarkably C yields, which were at highest 4.0% and 4.7%, respectively. As observed in Figure 3. synergy also existed in C formation over these binary systems. It should be noted that the maximal C yields over Ca-Ce, Ca-Cr, and CaMn systems are 3 - 4 times those observed with the single Pr or Tb oxide that is most effective for C formation (3). With regard to a binary catalyst reported earlier, a PbO-MgO system enhanced C yield in the oxidative coupling of CH4 with 0 in the coexistence of C 0 , but it readily lost the activity without 0 (11). Although a La 0 -ZnO system was also communicated to catalyze C formation from CH4 and C 0 (12), it was less active than the present binary catalysts. Since no mechanistic work to clarify not only the role of C 0 but the reaction mechanism has been carried out so far, the following sections focus on them. 3

2

2

3

2

2

2

3

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

2

2

2

2

2

2

2

2

2

3

2

2

2

Partial Pressure of C 0

2

The profiles for the reaction of CH4 alone with the Ca-Ce (4) or Ca-Cr system are illustrated in Figure 6, where CH4 diluted with He is used at P(CYU) of 30 kPa. In the absence of C 0 , H and CO were mainly produced with the molar ratio of approximately 2 by the reaction of CH4 with lattice oxygen atoms, and formation rates of H and COfinallyapproached to zero within 2 h. Figure 6 also revealed negligibly small amounts of C hydrocarbons even at the early stage of reaction, which indicates that the lattice oxygen of these oxide systems does not work as an oxidant for C formation. In the presence of C 0 , C formation proceeded readily. As shown in Figure 7, with the Ca-Ce catalyst, C yield became steady immediately after the start of reaction and did not change even when time on stream was prolonged to 10 h (6). Although it took 1 - 2 h for the yield to be steady with the Ca-Cr or Ca-Mn catalyst, their performances after the induction periods were sustainable, as with the Ca-Ce system. It is thus evident that oxygen species derived from C 0 play a crucial role in coupling reactions of CH4. The effect of P(C0 ) on CH4 conversion and product selectivity over the Ca-Ce catalyst is provided in Figure 8, where P(CKU) is kept constant (30 kPa) and inert He is used as a balance gas (6). The conversion increased steeply with 2

2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

92

Table L Surface Areas and Catalytic Performances of Binary Oxides of Alkaline Earth Metal and Ce Elements Surface area CH conversion C yield (%) (%) (m /g)

v

Alkaline earth metal M/Ce

2

4

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

2

}

Mg Ca Sr Ba

0.2 0.2 0.2 0.2

1.4 0.6 n.a* 0.5

Ca Ca"

0.5 0.5

0.4 1.4

11.6 4.9 3.1 2.8

0.2 2.5 1.6 1.3

5.8 5.4

3.2 0.7

2

Atomic ratio. ^Not analyzed. ^Physical mixture.

0

25

50

75

100

[Ca/(Ca+M)x100], at% Figure 4. Effect of catalyst composition on CH conversion with binary Ca-Cr and Ca-Mn catalysts at 850°C 4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

93

Ο

25

50

75

100

[Ca/(Ca+M) χ 100], at% Figure 5. C yield at 850°C over Ca-Cr and Ca-Mn catalysts with different compositions. 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

94

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

Ca-Cr "θ

ε 5—1

-co\

c ο

C

2

150

200

Time on stream /min Figure 6. Formation rates of products in the reaction at 850°C ofCH with Ca-Ce and Ca-Cr oxides.

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

alone

95

6

Ca-Mn

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

Ca-Ce ^Ca-Cr

ϋ

1 2

1 4

1 6

1 10

1 8

Time on stream /h

Figure 7. Effect of time on stream on C yield at 850°C over Ca based binary catalysts. 2

0

20

40

60

80

P(C0 ), kPa 2

Figure 8. Dependence ofCH conversion and product selectivity on partial pressure ofC0 over a Ca-Ce catalyst (Reproducedfrom reference 6 with permission from Academic Press). 4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

96 increasing P(C0 ) up to 10 kPa but leveled off beyond 20 kPa. This Langmuiriype curve indicates the involvement of C 0 chemisorption in the present reaction. With product selectivity, CO formation was predominant at low P(C0 ) below around 20 kPa, whereas C selectivity increased with increasing P(C0 ) and reached 70% at 70 kPa. As shown in Figure 9, the Ca-Cr and CaMn catalysts showed the similar CH4 conversion - F(C0 ) curves as with the Ca-Ce system, that is, the Langmuir-type relationship between the two. Furthermore, the Ca-Cr or Ca-Mn system provided almost the same dependency of C yield on P(C0 ); the yield increased almost linearly with P(C0 ). The observations in Figures 8 and 9 point out that high P(C0 ) is essentially needed for selective formation of C hydrocarbons. Since C 0 rather deactivates basic catalysts effective for the oxidative coupling of CH4 with 0 due to the strong adsorption ability, higher C selectivity under higher P(C0 ) is peculiar to the present reaction system. When time on stream was prolonged to 8 - 10 h, as shown in Figure 7, C yields were almost unchanged except for the early stage of reaction, irrespective of the kind of a binary oxide system. Such the stable performances mean that there is no catalyst deactivation by C 0 adsorption and carbon deposition. 2

2

2

2

2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

2

2

2

2

2

2

2

2

2

2

2

Chemisorption of C 0 and Catalyst State 2

In order to make clear the behavior of C 0 adsorption upon C formation, the three catalysts after reaction were subjected to the TPD runs. Figure 10 shows typical TPD profiles for the Ca-Ce system. The C0 -desorption peak appeared at 730°C under flowing pure He and shifted to 810, 850, and 910°C in a stream of C 0 with P(C0 ) of 10, 30 and 70 kPa, respectively (4,6). In other words, a pool of the C 0 chemisorbed existed on the catalyst in the process of C formation at 850°C under P(C0 ) of 70 kPa, whereas it almost disappeared under P(C0 ) of 10 kPa. The Ca-Cr and Ca-Mn catalysts after reaction also provided almost the same TPD profiles as in Figure 10. No significant desorption of C 0 from any fresh catalysts were observed. The comparison of Figures 8 and 10 suggests that the occurrence of C 0 chemisorption leads to high C selectivity. To ensure this point, the amount of C 0 chemisorbed at 850°C was estimated by integrating the peak area observed in Figure 10 between 850°C and 950°C. The estimated value is plotted in Figure 11 as a function of P(C0 ) used for the TPD run. The amounts chemisorbed on all of the binary catalysts showed almost the same dependencies on P(C0 ); the values were very small at around 10 kPa but larger at higher P(C0 ). It should be noted that such dependencies are quite similar as the C selectivity - P(C0 ) curves in Figure 8. It is thus likely that C 0 chemisorption plays a key role in selective formation of C hydrocarbons. 2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

97

P(CC>2), kPa Figure 9. Dependence ofCH conversion and C selectivity on partial pressure ofC0 over a Ca-Cr or Ca-Mn catalyst. 4

2

2

Figure 10. Profiles for C0 desorption during TPD measurements in different atmospheres of the Ca-Ce catalyst after the reaction of CH and C0 at 850°C. 2

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

98 The largest amounts observed in Figure 11 were 0.8 - 1.1 mmol, which were only 10 - 15% of Ca contents. Since any C 0 desorbed was not detectable with the used Ce0 , C r 0 , or M n 0 alone, Ca species on the surface layer of each binary catalyst must work as the sites for C 0 chemisorption. Table 2 summarizes the results of catalyst characterization. The XRD measurements showed that the Ca-Ce, Ca-Cr, or Ca-Mn catalyst before reaction existed mainly as Ca Cei_ 0 _ , CaCr0 , or CaMn0 , respectively. The former was stable even when time on stream was prolonged to 10 h, whereas the latter two were transformed to Ca(Cr0 ) and Cao^Mno 0, which were unchanged after 8 h. This transformation took place at the early stage of reaction. The XPS spectra revealed the partial or complete transformation of Ce , Cr , and M n to the corresponding reduced species, such as Ce , C r and M n . No XRD lines attributable to any carbonates were detectable with all of the three systems. As is seen in Figure 7, the Ca-Cr and Ca-Mn catalysts showed the induction periods before the steady performances were attained. The XRD and XPS observations mentioned above suggest that the periods are caused by the changes in oxidation states at bulk phases, since any induction period was not observed with the Ca-Ce catalyst on which surface modification only took place (Table II). 2

2

2

3

2

2

x

x

2

y

4

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

2

3

2

52

4+

3+

3+

6+

4 +

2+

Reaction Mechanisms Proposed The different mechanisms under low and high P(C0 ) may be proposed on the basis of the results mentioned above. At low P(C0 ), the main product from CH4 was CO (Figures 8 and 9), and the extent of C 0 chemisorption was very low (Figure 11), irrespective of the kind of a binary catalyst. Since CH4 reacted readily with the lattice oxygen to form CO and H (Figure 6), this reaction may proceed predominantly at low P(C0 ) through the following scheme: 2

2

2

2

2

CH4 + CaMO —• CO + 2H + CaMO^ x

2

CaMOx.i + C 0 -> CO + CaMO 2

x

(3)

(4)

where M denotes Ce, Cr, or Mn element. At high P(C0 ) above 60 kPa, on the other hand, selective formation of C hydrocarbons took place, and C selectivity exceeded 60% on all of the binary catalysts examined (Figures 8 and 9). Almost the same dependencies of C selectivity and C 0 chemisorption on P(C0 ), observed in Figures 8, 9, and 11, 2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

99

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

1.2

P(C0 ), kPa 2

Figure 11. Amounts of the C0 chemisorbed on Ca-based binary catalysts after reaction under different partial pressures ofC0 . 2

2

Table Π· Species Identified by XRD and XPS Measurements of Ca Based Binary Catalysts Catalyst

Before reaction XRD XPS 2;

Ca-Ce (0.5) Ca-Cr (l.0) Ca-Mn (1.0) 2)

1}

C^CeuA-y, CaCr0 CaMnQ

3)

Ca0

6 +

4

2;

3

2;

At a steady state. Atomic ratio.

4+

Ce Cr Mn

3;

}

After reaction XRD XPS C a ^ e ^ C ^ y , Ca0 Ca(Cr0 ) , C a O Cao.48Mn . 0 j;

2

4 +

2

0

52

3;

4+

C e , Ce" Cr Mn

With very small XRD intensities.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

3 +

2 +

100

indicate that C formation involves the chemisorption process. The following scheme may be proposed: 2

C 0 -> C 0 (a) -> CO + O*

(5)

CH4 + O* —• CH « + OH-

(6)

2CH - -> C H6 -+ C H4

(7)

2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

3

3

2

2

where C0 (a) and O* designate chemisorbed C 0 and active oxygen species, respectively. The C 0 in feed gas first adsorbs on basic C a sites, and the reduced sites, such as Ce , C r and M n , then activate the C 0 chemisorbed to provide CO and active oxygen species adsorbed on the corresponding oxidized sites (6). These sites appear to be near Ca , since the three catalysts existed as the composite oxides (Table 2). According to equation (6), CH4 reacts with the oxygen species to form methyl radicals, which subsequently undergo coupling reactions. In the former process, the regeneration of Ce, Cr, and Mn sites take place (6). Since molar ratios of C H 4 / C H observed were almost independent on P(C0 ), C 0 seems to be hardly involved in the dehydrogenation of C H to C H4. Thus, C formation at a steady state proceeds probably through a cycle mechanism between the reduced and oxidized site of Ce, Cr, or Mn species. 2

2

2+

2

3+

3+

2+

2

2+

2

2

2

2

6

2

6

2

2

Conclusions Binary catalysts of CaO and Ce, Cr, or Mn oxide show remarkable synergistic effects on formation of C hydrocarbons from CH4 and C 0 at 850°C, and the maximal C yields are 30 - 70 times that over each component. As partial pressure of C 0 increases, C selectivity over all catalysts increases and reaches 65 - 75% at 70 kPa. Their performances are stable during 8 - 10 h reaction. The TPD, XRD, and XPS measurements strongly suggest that C 0 chemisorption on Ca sites and subsequent activation on neighboring Ce, Cr, or Mn sites provide active oxygen species that play a key role for selective formation of C hydrocarbons. 2

2

2

2

2

2

2

Acknowledgement The present work was supported in part by a Grant-in-Aid for Scientific Research (B) from the Ministry of Education, Science, Sports and Culture, Japan (No. 10555275).

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

101

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch006

References 1. Asami, Κ.; Fujita, T.; Kusakabe, Κ.; Nishiyama.Y.; Ohtsuka, Y. Appl. Catal. A: General 1995, 126, 245 2. Asami, K.; K., Kusakabe, K.; Ashi, N.; Ohtsuka, Y. Stud. Surf. Sci. Catal. 1997, 107, 279. 3. Asami, K.; Kusakabe, K.; Ashi, N., Ohtsuka, Y. Appl. Catal. A: General 1997, 156, 43. 4. Wang, Y.; Takahashi, Y.; Ohtsuka, Y. Appl. Catal. A: General 1998, 172, 203. 5. Wang, Y.; Takahashi, Y.; Ohtsuka, Y. Chem. Lett. 1998, 1209. 6. Wang, Y.; Takahashi, Y.; Ohtsuka, Y. J. Catal. 1999, 186, 160. 7. Wang, Y.; Ohtsuka, Y. J. Catal. 2000, 192, 252. 8. Lunsford, J.H. Angew. Chem., Int. Ed. Engl. 1995, 34, 970. 9. Ohtsuka, Y.; Asami,K.;Wang, Y. J. Chin. Inst. Chem. Engrs. 1999, 30, 439. 10. Kuo, J.W.; Kresge, C.T.; Palermo, R.E. Catal. Today 1989, 4, 470. 11. Nishiyama, N.; Aika, K. J. Catal. 1990, 122, 346. 12. Chen,C.;Xu, Y.; Li, G.; Guo, X.; Catal. Lett. 1996, 42, 149.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

Chapter 7

Copolymerization of Carbon Dioxide, Propylene Oxide, and Cyclohexene Oxide by an Yttrium-Metal Coordination Catalyst System Chung-Sung Tan, Char-Fu Chang, and Tsung-Ju Hsu Department of Chemical Engineering, National Tsing Hua University, Hsinchu, Taiwan 30043, Republic of China

The catalyst system consisting of Y(CF CO ) (I), diethylzinc (II), and glycerine (III) in the solvent of 1,3-dioxolane was found to be very effective to carry out copolymerization of carbon dioxide, propylene oxide (PO), and cyclohexene oxide (CHO). At 353K, 27.2 atm., and a proper combination of the components I to III, the yield and weight-average molecular weight could be as high as of 7948 g/(mol of Y)/h and 2.53 x 10 , respectively, for a 12 h operation. The IR and NMR spectra indicated that the resulting copolymer was an alternating polyethercarbonate. From the DSC analysis, it is observed that the copolymer resulting from an equal or a higher feed ratio of PO to CHO possessed two glass transition temperatures (T ), otherwise only one T . The temperature at which 10 % weight loss occurred (T ) of the resulting polyethercarbonate was also measured in this study. Both T and T were found to locate in a range between those of the copolymers generatedfromCO andPO,and CO and CHO. 3

2 3

5

g

g

10

g

10

2

102

2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

103

The copolymerization of C 0 with epoxides to result in polyethercarbonate is one of the possible means for utilization of C 0 . Besides, this means can avoid using more toxic compound such as phosgene in synthesis. But due to C 0 is a relatively unreactive compound, an effective catalyst is generally required. For production of an alternating polyethercarbonate using PO or C H O as monomer, the catalyst systems consisting of dialkylzinc and a compound having two active hydrogens (1-5) as well as other synthesized zinc-based organometallic catalysts (6-12) have been extensively studied. Using these kinds of catalyst systems, a long reaction time is usually required and the yield needs to be improved. A ternary rare-earth-metal coordination system Y(P 04)3 -Al(i-Bu) -glycerine was reported recently to carry out the copolymerization of C 0 and P O (13). Despite that a random polyethercarbonate was produced, a higher molecular weight with a narrower molecular weight distribution and a higher yield within a shorter reaction period could be obtained compared to the binary organometallic catalyst systems. When the ternary rare-earth-metal coordination system containing Y ( C F C 0 ) 3 , Zn(Et) , and glycerine was used to proceed the copolymerization of C 0 and PO, an alternating polycarbonate with a carbonate content of 95.6 % instead of a random polyethercarbonate could be generated (14). In addition to carbonate content, a much higher yield was also observed using this reaction catalyst system. For the copolymerization of C 0 and CHO, this ternary yttriummetal coordination catalyst system was also proved to be effective as well (15). Though a high molecular weight of polyethercarbonate can be generated either by the copolymerization of C 0 with PO or C H O using yttrium-metal coordination catalyst system, thermal properties of the resulting copolymers are quite different, for example, T are about 308 Κ and 385 K , and T are about 463 Κ and 555 Κ for the copolymers generated from C 0 and PO and C 0 and C H O , respectively. While the thermal properties of the resulting copolymer can be improved using C H O instead of PO as monomer, the mechanical properties were found to reduce. It seems to be rational to use both PO and C H O as monomer in copolymerization reaction to generate a polyethercarbonate with the desired thermal and mechanical properties. The objective of this paper is therefore to systematically study the copolymerization of C 0 , PO, and C H O catalyzed by a ternary yttrium-metal coordination catalyst system. 2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

2

2

3

2

3

2

2

2

2

2

g

i 0

2

2

2

Experimental Propylene oxide of a purity of 99.5 % (Janssen Chimica) and cyclohexene oxide of a purity of 98 % (Lancaster Synthesis) were treated by vacuum distillation before use. Diethylzinc, glycerine, and all the solvents, such as nhexane, toluene, D M S O , 1,3-dioxolane, and THF, were of analytical reagent and used without further purification. C 0 with a purity of 99.99 % (Air Product) 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

104

was used as received. Yttrium trifluoroacetate, yttrium acetate, yttrium 2ethylhexanonate, yttrium acetylacetonate, and yttrium nitrate purchased from Aldrich Chemical Co. were heated in vacuum at 353 Κ for 40 h before use. The catalyst system containing yttrium compound, Zn(Et) , and glycerine was prepared in an atmosphere of purified argon. Glycerine was added dropwise to a solution of Zn(Et) in solvent at room temperature. After ethane gas evolution had ceased, the solution containing white powders resulting from the reaction between Zn(Et) and glycerine was heated at 333 Κ for 2 h. This solution was then added to an autoclave that contained a known amount of yttrium compound. The resultant catalyst was stirred at 353 Κ for 1 h prior to C 0 , PO, and C H O were introduced. Copolymerization of C 0 , PO, and C H O was carried out in a 300 ml autoclave that was equipped with a magnetic stirrer (Autoclave Engineers Inc.). The spinning speed was kept at 1000 rpm. After a certain period of reaction time, the pressure in the autoclave was allowed to reduce to atmosphere. To collect the resulting copolymer in the autoclave, THF was introduced into the autoclave first to dissolve the copolymer and then the solution was added by an aqueous methanol solution to cause a precipitation. The copolymer reprecipitated was dried in vacuum at 303 Κ for 8 h before analysis. To determine the structure and the composition of the copolymers generated, the Ή and C N M R spectra were measured by a Bruker AM-300 N M R spectrometer and the IR spectra by a Jasco J-0085 spectrometer. The molecular weights were determined by a gel permeation chromatography (Waters 150-CV). The glass transition temperature and T were measured by DSC (duPont 2900) and T G A (duPont951). 2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

2

2

2

1 3

J 0

Results and Discussion In the beginning of the study, screening of solvents was first performed. Table I indicates that a less polar solvent- n-hexane and a more polar solventD M S O were not a good candidate to make polyethercarbonate from C 0 , PO, and C H O . Regarding yield and molecular weight distribution, 1,3-dioxolane was found to be superior to toluene. At least 40 % increase in yield and a narrower molecular weight distribution of the resulting copolymer using 1,3-dioxolane were observed. These results indicate that a proper choice of solvent is essential in this kind of copolymerization. Since 1,3-dioxolane showed its excellence, it was used as solvent in the subsequent runs. 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

105

Table I. Effect of Solvent on Copolymerization of C 0 , PO, and C H O Using the Catalyst System Y(CF C0 )3-Zn(Et)2-Glyeerine* 2

3

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

Solvent

yield, z/(molofY)/h

2

Mw χ

Μη χ 10'

4

5

ισ

Mw/Mn

1,3-Dioxolane

7948

5.5

2.53

4.63

Toluene

5152

3.4

2.13

6.20

n-Hexane

Trace

...

...

—

DMSO

Trace

...

...

—

* I = 0.0004 mol; II = 0.008 mol; III = 0.004 mol; Τ = 353 Κ; Ρ = 27.2 atm; time = 12 h

Table II shows the experimental results using different yttrium coordination catalyst systems in 1,3-dioxolane at 353 Κ and 27.2 atm. Each experiment was performed in duplicate, the difference in yield was observed to be always less than 7.0 %. It is seenfromTable II that the catalyst system Y(CF C0 )3-Zn(Et) - glycerine provided the highest yield. It might be due to that fluorine induced more positive charge of yttrium, in a consequence, the coordination catalyst could adsorb C 0 more easily. This catalyst system was also proved to be the most appropriate one for the copolymerization of C 0 with PO or CHO (14, 15). The IR, H - and C - NMR spectra of the resulting copolymer, shown in Figures 1 to 3, respectively, indicate that the copolymer was an alternating polyethercarbonate and the copolymerization reaction took place according to the following scheme, 3

2

2

2

2

l

1 3

X7

O

Œ

3

Ο

\^

J

Since the poly(propylene oxide) and poly(cyclohexene oxide) contents in the resulting copolymer were small, sorption rate of C 0 on the present catalyst was believed to be fast. 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

106

Table IL Copolymerization of C 0 , PO, and C H O by Different Catalyst Systems* 2

catalyst system

Yield, Μη

Y(CF C0 )

2

Glycerine

7948

Zn(Et)

2

Glycerine

Trace

Y(CH COCH= Zn(Et) C(0-)CH )

2

Glycerine

Y(CH (CH ) C Zn(Et) Glycerine H(C H )C0 )

2

3

Y(CH C0 ) 3

2

3

3

3

3

2

2

5

3

3

Mw

χ 70° Mw/Mn

5.5

2.53

4.63

5644

6.6

2.63

3.97

6069

7.0

2.52

3.58

6663

7.5

2.85

3.83

3

3

2

Y(N0 )

χ lCT

g/(molofY)/h

Zn(Et)

3

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

(III)

(Π)

(I)

2

3

Zn(Et)

2

Glycerine

I = 0.0004 mol; II = 0.008 mol; III = 0.004 mol; Τ = 353 Κ; Ρ = 27.2 atm; time = 12 h

CM

Figure

1. IR spectrum of the resulting

copolymer.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

107

a Ο

b

a b'

r

CH3

CH3

C

c

T

[Hid

r

1

ι

,

Γ.Η'.ΗΊd

R

Η

T

c,e,f c',e ,f f

Η

fi Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

Solvent Peak

T

f

a ,b ,c

9.0

8.0

7.0

6.0

5.0

4.0

f

3.0

2.0

1.0

0.0

PPM

Figure 2. ^-NMR spectrum of the resulting copolymer.

180

170

160

150

140

130

120

110

100 90 PPM

80

70

60

50

40

30

20

13

Figure 3. C-NMR spectrum of the resulting copolymer.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

10

108

When the molar ratios of II/I and III/II were maintained at 20 and 0.5, respectively, and the pressure was fixed at 27.2 atm., Table III shows that the yield increased with increasing temperature in a lower temperature range, reached a maximum at about 353 K , and then dropped as the temperature further increased. The existence of an optimum temperature in the copolymerization of C 0 , PO, and C H O is also observed in the copolymerization of C 0 and PO (14)as well as C 0 and C H O (15). According to the transition state theory (16), high-pressure operations are usually more favorable to polymerization reactions. For this copolymerization reaction, the yield was indeed enhanced with pressure as the pressure was less than 27.2 atm., but after that pressure was reached, the yield decreased with increasing pressure, shown in Table IV. One of the possible reasons for the decrease in yield is due to the decrease in solubility of copolymer resulted from the swelling of the solvent caused by the dissolution of C 0 . 2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

2

2

Table III. Effect of Temperature on Copolymerization of C 0 , PO, and CHO* 2

T κ t

yield, g/(molofY)/h

4

Μη χ W

5

Mw χ 10~

Mw/Mn

323

3475

7.4

5.02

6.94

333

6027

6.3

4.34

6.91

343

6792

6.0

3.78

6.33

353

7948

5.5

2.53

4.63

363

6046

4.2

2.46

5.88

373

5683

3.5

1.76

5.08

* I = 0.0004 mol; II = 0.008 mol; III = 0.004 mol; Ρ =27.2 atm.; time = 12 h :

Table IV. Effect of Pressure on Copolymerization of CO., PO, and CHO* P, atm. yield, z/(mol ofY)/h Μη χ 10' Mw x JO' Mw/Mn 4

20.4

6394

23.8 27.2 30.6 34.0 40.8 47.6

6717 7948 7710 7604 7210 7069

5.7 6.0 5.5 5.3 5.3 5.9 5.3

5

2.66 3.81 2.53 2.95 3.15 3.29 3.37

* I = 0.0004 mol; II = 0.008 mol; III = 0.004 mol; Τ = 353 Κ; time = 12 h

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

4.64 6.34 4.63 5.58 5.96 5.55 6.35

109

Tan and Hsu (14) pointed out that the component ratios in a rare-earth-metal coordination catalyst system have significant effects on yield and molecular weight in copolymerization of C 0 and PO, the same conclusion can be drawn here. From the experimental results over wide ranges of the molar ratios of II to I and III to II, it was observed that the II/I and III/II ratios of 20 and 0.5, respectively, provided the highest yield though not the largest molecular weight and the best Mw/Mn. It is seen from Table V that only one T was observed when the feed contained more C H O than PO. On the other hand, two glass transition temperatures existed. This observation might provide two aspects: 1). a random polyethercarbonate or a block-polyethercarbonate could be produced by changing the monomer ratio in feed, 2). the ring-opening rate of C H O is faster than that of PO. At any ratio of PO to CHO, T of the resulting copolymers are all larger than those of the copolymer generated from C 0 and PO but less than those of the copolymer generated from C 0 and CHO. In addition, the T of the resulting copolymers were found to locate in a range between those of the copolymers generated by using PO or C H O as monomer. 2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

g

g

2

2

1 0

Conclusion Polycarbonate could be produced effectively from the copolymerization of carbon dioxide, propylene oxide, and cyclohexene oxide in 1,3-dioxolane with a catalyst system consisting of Y ( C F C 0 ) , diethylzinc, and glycerine. A t a proper combination of temperature, pressure, and molar ratios of the catalyst components, a very high yield up to 7948 g/(mol of Y)/h and a molecular weight 3

2

3

Table V. Effect of Monomer ratio on Yield and Molecular Weight* PO ml 10 10 10 15 20 20 30 40 30

CHO ml 20 30 40 15 20 10 10 10 20

Yield

* 33.24 40.32 47.75 38.15 40.67 28.59 35.56 37.51 44.78

% 76.72 68.94 64.82 90.34 72.23 69.50 65.71 55.90 64.51

4

5

Μη x Iff Mw x Iff 9.3 6.4 8.3 5.5 5.2 5.3 3.5 6.5 7.1

3.72 2.08 3.88 2.53 1.75 1.66 1.06 2.29 2.55

Mw/Mn

T ,K

Τι* Κ

4.02 3.23 4.67 4.63 3.40 3.13 3.03 3.52 3.58

343 371 390 313,370 319,371 319,361 320,357 319,368 321,371

549 545 550 543 552 533 544 528 540

g

* I = 0.0004 mol; II = 0.008 mol; III = 0.004 mol; Ρ = 27.2 atm.; Τ = 353 Κ; time = 12 h

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

110 5

as high as of 2.5 χ 10 were observed to obtain in a 12 h run in a batch-mode operation. A copolymer containing random distribution of PO carbonate and CHO carbonate or a block-polyethercarbonate could be generated by varying the volume ratio of the monomers PO and CHO in feed. When the volume of CHO was greater than that of PO, only one T of the resulting copolymer existed, otherwise, two glass transition temperatures existed. All the T and T of the present copolymer were found to be present in a range between those of the copolymer generated from C 0 and PO and C 0 and CHO. This indicates that the thermal properties of the resulting polyethercarbonate can be modified by introducing two epoxides with different chemical structure. g

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

g

2

1 0

2

Acknowledgment Financial supportfromthe National Science Council of Republic of China, Grant NSC89-2214-E007^009, is gratefully acknowledged.

References 1. 2. 3.

Inoue, S.; Koinuma, H.; Tsuruta, T., Makromol Chem., 1969, 130, 210. Yoshida, Y.; Inoue, S., Polym. J., 1980, 12, 763. Kuran, W.; Pasynkiewicz, S.; Skupinska, J., Makromol. Chem., 1977, 178, 2149. 4. Hino, Y.; Yoshida, Y.; Inoue, S., Polym. J., 1984, 16, 159. 5. Nishimura, M.; Kasai, M.; Tsuchida, E., Makromol. Chem., 1978, 179, 1913. 6. Inoue, S.; Koinuma, H.; Yokoo, Y.; Tsuruta, T., Makromol. Chem., 1971, 143, 97. 7. Darensbourg, D. J.; Holtcamp, M. W., Macromolecules, 1995, 28, 7577. 8. Darensbourg, D. J.; Zimmer, M. S., Macromolecules, 1999, 32, 2137. 9. Sarbu, T.; Beckman, E. J., Macromolecules, 1999, 32, 6904. 10. Super, M.; Berluche, E.; Costello, C.; Beckman, E., Macromolecules, 1997, 30, 368. 11. Ree, M.; Bae, J. Y.; Jung, J. H.; Shin, T. J., J. Polym. Sci. Polym. Chem., 1999, 37, 1863.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

111

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch007

12. Darensbourg, D. J.; Holtcamp, M. W.; Struck, G. E.; Zimmer, M . S.; Niezgoda, S. Α.; Rainey, P.; Robertson, J. B.; Draper, J. D.; Reibenspies, J. H., J. Am. Chem. Soc., 1999, 121, 107. 13. Chen, X.; Shen, Z.; Zhang, Y., Macromolecules, 1991, 24, 5305. 14. Tan, C. S.; Hsu, T. J., Macromolecules, 1997, 30, 3147. 15. Hsu, T. J.; Tan, C. S., submitted to J. Polymer Sci., 2000. 16. Savage, P. E.; Gopalan, S.; Mizan, T. I.; Martino, C. J.; Brock, E. C., AIChE J., 1995, 41, 1723.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 8

The Role of C O for the Gas-Phase O Oxidation of Alkylaromatics to Aldehydes Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

2

2

Jin S. Yoo Samsung Chemical Company, 2315 Mast Court, Flossmoor,IL60422

Abstract: A synergistic cooperation between the guest moiety and the host matrix in the CVD Fe/Mo/DBH catalyst gave rise to a novel catalytic activity for the gas-phase O oxidation of alkylaromatics. Para-selective property and activation of C O to function as a co-oxidant are two key outstanding attributes exhibited by the catalyst. The oxidation mechanism was remarkably altered in the co-presence of C O in the O feed stream. The function of C O not only enhanced the rate of reaction, but also improved the selectivity toward the aldehyde/oxygenate products. The peroxocarbonate complex has been postulated as an active catalytic species, which promoted the initial abstraction of hydrogen atom from the alkyl group in a much more effective manner than the usual O oxidation. 2

2

2

2

2

2

During the last two decades, carbon dioxide has been attracting increasingly more attention worldwide for the green chemistry. Many homogeneous and heterogeneous catalysts have been reported for activation of the CO2 molecule and its usage as an abundant inexpensive Ci-source for the chemical industry (1-10). The notable advances in its utilities include a oneoxygen transfer oxidant and a reaction modulating agent (11-14) for the synthesis of valuable commodity chemicals, fine and specialty chemicals containing the functional groups such as olefin (15-23), aldehyde, ketone, acid (24,25), epoxide, ester (26), carbonate (2,27), lactone, alcohol (28-31), carbamate (32-33), amide and other useful oxygenates (34). in light of the unique properties of CO2 exhibited with a variety of catalysts, it deems

112

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

113 appropriate to define the role of CO2 as an oxidant more clearly for the organic syntheses as well as for green chemistry. In the course of developing a new process for the selective synthesis of terephthaldehyde by the gas-phase 0 oxidation of p-xylene over C V D Fe/Mo/DBH during late 1980s, it was unexpectedly discovered mat the oxidation mechanism was drastically altered in the co-presence of C 0 in the 0 feed stream. This finding prompted us to continue to study the role of CO2 for the gas-phase 0 oxidation of various alkylaromatics to the corresponding aldehydes, oxygenates and, in some cases, olefins via oxidative dehydrogenation. 2

2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

2

Results and Discussion The C V D Fe/Mo/DBH catalyst was prepared from FeCl , MoOzCk or M0CI5 and partially deboronated borosilicate molecular sieve known as HAMS1B-3 (DBH) by the chemical vapor deposition technique based on a new catalyst concept in the late 1980s (1,35). The CVD Fe/Mo/DBH catalyst consisted of two guest phases of ferric molybdate, Fe (Mo0 ) , and M0O3 (a minor fraction) uniformly stabilized as nano-size particles within the main channel structure of the host MFI structure of the DBH matrix (36,37). A synergistic cooperation between guest nano-particles and the host matrix brought about novel catalytic activities, namely, para-selective property and a unique function of activating the CO2 molecule to function as a co-oxidant for the gas-phase 02 oxidation of a variety of alkylaromatics. 3

2

4

3

Oxidation of alkylaromatics over CVD Fe/Mo/DBH Xylene isomers, pseudocumene, durene, mesitylne, substituted toluenes, ethylbenzene and substituted ethylbenzene derivatives were among a variety of alkylarornatics studied in this work. Besides an unexpected property of activating the C Q 2 molecule (38), the catalyst also exhibited other interesting properties such as para-selectivity and catalyst stability for the oxidation reactions to produce aldehydes and/or oxygenates (39-46) as well as dehydrogenation of alkylbenzenes such as ethylbenzene, diethylbenzene, isopropyltoluene and ethyltoluene and alkanes to yield the corresponsing olefins (39,43, 44).

Oxidation of p-Xylene over CVD Fe/Mo/DBH In a typical gas-phase O2 oxidation of p-xylene over the C V D Fe/Mo/DBH catalyst, terephthaldehyde and p-tolualdehyde were produced in a high yield despite the extremely high surface area of the catalyst (150-300 m /g). The co2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

114

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

presence of CQ> not only enhanced the rate of oxidation, bit also increased the selectivity and created a unique activity to alter the product distribution. The oxidation temperature was remarkably lowered as much as 150°C~200°C in the co-presence of CCh than that required to attain the same conversion of the pxylene with O2 alone.

The experimental results on the p-xylene oxidation are summarized in Table I. The author has postulated two different active catalytic sites, "pair" and "single" sites, for the formation of terephthaldehyde (I) and p-tolualdehyde (II) elsewhere (1,27); The p-xylene molecule was aligned on the "pair sites" in a parallel fashion for activation, and one Η-atom from each methyl group was simultaneously abstracted in a concerted manner to produce the stable diradical, which led to the aldehyde products via the peroxide intermediate. On the "single" sites, p-xylene was activated in the stand-on manner, and one H was abstractedfromthe activated methyl group to yield a single radical, which led to produce r>tolualdehyde (19). Further oxidation of p-tolualdehyde produced ptoluic acid instead of terephthaldehyde. Terephthaldehyde was not produced by the consecutive oxidation of p-tolualdehye. Instead, these were produced on two different active centers. It became possible to control the molar ratio of terephthaldehyde to ptolualdehyde in the product (Ι:Π) in a wide range of 4:1 to 1:4 by adjusting factors such as the loading of the catalyst component, the acidity of the DBH host matrix, the atomic ratio of Fe/Mo in the guest moiety, the reaction temperature, and the composition of the feed. Among these factors, the copresence of CO2 in the O2 feed stream appears to be the most dominant actor. It mustftmdamentallyalter the reaction mechanism by affecting not only the oxidation rate, but also controlling the product yield These conclusions are also supported by the results plotted in Figure 1. The

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

115

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

250 275 300

350

Temperature C

325

375

400

425

2

Figure 1. Effect ofC0 on gas-phase oxidation of p-xylene over CVD Fe/Mo/DBH as a function of reaction temperature. Catalyst composition: 6.80% Mo, 1.88% Fe, 54 ppm B, the atomic ratio of Mo/F e was alteredfrom 2.11 for fresh non-activated system to 1.89 for the catalyst activated by prolonged air calcination. Reproduced with permission from reference 38. Elsevier 1993.

CL

X

>

C Ο

v.

Ό)

c ο

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

117 conversion of p-xylene was remarkably increased in the co-presence of CO2 in the O2 stream in comparison to that obtained in O2 alone at the oxidation temperatures, 250~450°C. Under the anaerobic condition in the presence of CQ2 without Q2 in the feed stream (see Figure 2), the conversion of p-xylene was drastically reduced to a level of 5-6% and maintained at that level throughout a 45 hour run.

Table L Gas-phase 0 oxidation of p-xylene over CVD Fe/Mo/DBH Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

2

Reaction temperature: 350°C, the concentration of p-xylene in the 0 feed stream: 0.1 mol% 2

Feed gas

02 02+C02 CU2

p-xylene Conversion % | 41.2 65.5 5

Product selectivity mol% II byproduct others 8

23.5 33.6 14

50.2 40.9 69

5.2 5.4 17

0

4.8 8.7

CO

CÔJ

0.6 3.4

15.5 8.0

C

Lterephthaldehyde ïï:p-tolualdehyde a:hydrocarbon byproducts produced by the side reactions catalyzed by the acidic site of the DBH matrix b:fwther oxidized products c:most likely produced from the C 0 feed instead of burning. 2

The selectivity of terephthaldehyde declined rapidly, and became nil in 10 hours on stream while the selectivity of p-tolualdeyde kept at an approximately 30% level. When the deactivated catalyst was calcined in the air, the selectivity to terephthaldehyde was restored again at a level of 10%, but it deactivated very quickly. However, the results shown in Figure 3 indicate that terephthaldehyde was not produced and even the selectivity of p-tolualdehyde, a sole aldehyde product, decreased extremely fast and became negligible in a flow of N without CO2 within a 5 hour period. These results suggest that CO2 actually participates in the formation of active oxygen species even under the anaerobic condition although die process proceeds extremely slolowly. The promoting phenomenon of CO2 and para-selective activity were also observed with polymethylbenznes such as pseudocumene, durene, 2,6dimethylnaphthalene and 4,4'-dimethylbiphenyl (41), substituted toluene derivatives (42), other types of alkylaromatics (43) with the CVD Fe/Mo/DBH catalyst. The same promoting function of C 0 was also found for the gas-phase O2 oxidation of alkane over both CVD Fe/Mo/DBH and CVD Zr/W/DBH (43), 2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002. 2

Figure 2. Oxidation ofp-xylene in C0 alone over CVD Fe/Mo/DBHconversion and selectivity vs hours on stream

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002. 2

Figure 3. Reaction of p-xylene in N over CVD Fe/Mo/DBH TMBPM: trimethylbiphenylmethane

H o u r s on s t r e a m

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

120

Chemical modification of CVD Fe/Mo/DBH with silica deposition, Si/Fe/Mo/DBH Attempts to characterize the active catalytic sites were made by the chemical modification techniques. The "single" site of the CVD Fe/Mo/DBH catalyst was more predominantly poisoned by depositing silica using Si(OMe) via the chemical deposition method. As shown in Table Π. the molar ratio of terephthaldehyde to p-tolualdehyde in the product remarkably increased from 0.51:1 with the unmodified system to 0.98:1 over die silica modified catalyst, CVD Si/Fe/Mo/DBH (39) for the oxidation of p-xylene. The catalytic activity was about the same (51-55% conversion) for both modified and unmodified original catalyst.

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

4

Table E L Oxidation of p-xylene over silica modified catalyst, CVD Si/Fe/Mo/DBH

Temperature °C p-Xylene conversion % Selectivity (mol%) Terephthaldehyde (I) p-Tolualdehyde (Π) Molar ratio of I : II

CVD Si/Fe/Mo/DBH 350 51.0

CVD Fe/Mo/DBH 350 54.6

32.2 32.7 0.98 : 1

24.9 48.4 0.51: 1

These results indicate that the selectivity to p-tolualdehyde was noticeably decreased by preferential poisoning of the "single" sites over the "pair" sites. It also suggests that the "single" sites may be mainly located on the exterior surface of the catalyst since the size of the Si(OMe) is too big to get an access to the main channel structure of the host MFI matrix in the silica deposition procedure. It is likely that silica must be deposited on the exterior surface, while the guest Fe (Mo0 )3 moiety inside the channel remains intact. Thus one could conclude that the "pair" sites are primarily located inside the channel: the limited space inside the channel structure of CVD Si/Fe/Mo/DBH may force the p-xylene molecule to align itself on the "pair" site in a parallel fashion for the 4

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

121 concerted abstraction of one hydrogen atoms from two methyl groups to generate a stable diradical, · Ο Ϊ ^ 6 Η 4 - α ΐ · , which leads to form terephthaldehyde via the peroxide radical intermediate, ^ C ^ - C ^ - C I t C V . 2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

Chemical modification of CVD Fe/Mo/DBH with Ag-doping, Ag/Fe/Mo/DBH Silver was doped at a level of 760 ppm on CVD Fe/Mo/DBH by using an aqueous solution of Ag(N0 )3 via the impregnation method to prepare the Ag/Fe/Mo/DBH catalyst. The original "pair" sites on die Fe/Mo/DBH catalyst were effectively poisoned by this procedure, and thus p-tolualdehyde became almost a sole product (45). 3

Table HI- Oxidation of p-xylene over Ag/Fe/Mo/DBH Catalyst Ag/Fe/Mo/DBH CVD Fe/Mo/DBH 256 169 Surface area m /g 350 350 Temperature °C 2 1 2 1 Feed type 41.2 65.6 13.6 32.3 Conversion % Product selectivity mol% 50.2 60.5 90.7 79.3 p-Tolualdehyde 23.5 33.6 2.6 1.0 Terephthaldehyde 2.4 2.7 0.4 0.4 Benzaldehyde 3.2 11.5 0 0 Maleic anhydride 3.5 4.8 1.2 0.8 Toluene Trimethylbiphenyl0.6 0.1 1.7 methane(TMBPM) 0 3.2 0.6 4.9 0 CO 15.6 18.5 COz a:excludmg Cu ; teed 1:U.1% p-xylene, 1.0% 0 , 1.0% N in He; feed 2:0.1% p-xylene, 1.0% 0 ,1.0% N in CQ 2

a

a

2

2

2

2

2

2

The catalytic activity kept intact while the CO2 promoting function still remained with Ag/Fe/Mo/DBH, as shown in Table III. However, another separate run with Si/Fe/Mo/DBH showed that the promoting function of C 0 was suppressed by doping the silica on the CVD Fe/Mo/DBH catalyst 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

122

Oxidation of Ethylbenzene Derivatives

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

In the course of pursuing the selective oxidation of the xylene isomers containing ethylbenzene over the CVD Fe/Mo/DBH catalyst, we found that the catalyst exhibited an excellent activity to produce styrene as an initial product via the oxidative dehydrogenation reaction (44). If the alkyl groups in the alkylaromatics are ethyl or higher, dehydrogenation of the alkyl group to produce alkenylaromatics becomes the primary reaction with the catalyst under oxidation conditions similar to those for polymethylaromatics. The similar promoting effect of C 0 was also observed with ethylbenzene derivatives, in particular, ethylbenzene and p-ethyltoluene (see Table IV). Ethylbenzene was converted to styrene and the ethyl group in p-ethyltoluene was preferentially dehydrogenated to produce p-methylstyrene over the methyl group at the initial stage of the reaction. As soon as these styrene products were formed, they were very rapidly oxidized further to the corresponding aldehyde and oxygenates. 2

Table IV, Oxidation of p-ethylbenzene and Styrene in C 0 plus 0 over CVD Fe/Mo/DBH 2

Substrate Temperature °C Feed type Conversion %

p-Ethyltoluene 375 1 2 39.0 60.0 Product selectivity % p-Methylstyrene 7.3 6.4 p-Tolualdehyde 17.2 20.0 p-methylacetophpenone 6.9 6.7 p-Toluicacid 23.0 34.8 Terephthaldehyde 0.6 1.2 Benzaldehyde Phenylacetaldehyde Acetophenone Benzoic acid CO 14.8 22.0

2

Styrene 300 1 2 34.5 26.5

a

19.9 1.9 6.5 57.3 6.6

b

21.T 2.5 7.7 60.5 5.6

b

a:data exclude CO2 (all data in the column), bilikely to be formedfromC 0 instead of burning source, Feed 1:0 alone, feed 2: C 0 plus 0 2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

123 On the contrary, when styrene was subjected to the oxidation reaction over the same catalyst under similar conditions, the co-presence of CO2 in the Q2 stream hampered the oxidatioa In short, the C 0 molecule participates in die abstraction of H from the methyl or alkyl group in die gas-phase O2 oxidation to produce styrene at the initial stage, while it remained as an inert diluent for the subsequent oxidation of the styrene to other oxygenates in the second stage. The CO2 plays a key role in the initial abstraction of H from methyl or alkyl group, but it becomes an inert gas for the oxidation of unsaturated substrates such as styrene and its derivatives. In this respect, die C V D Fe/Mo/DBH catalyst behaves quite differently from the homogeneous catalysts such as [Rh(diphos)Cl].Q^ (11,12) and Rh(bipy)(C H4)Cl (11), which exhibited a remarkable role as a one-oxygen transfer oxidant as well as a modulating agent for the oxidation of styrene and tetrahydrofuran.

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

2

2

The Oxidation Mechanism A synergistic interaction of the catalytic species with DBH In order to understand the synergistic interaction between the guest catalytic moiety of Fe Mo0 )3 and the host DBH matrix having an MFI structure stated in thefrontsection of this paper, the catalytic behavior of the DBH matrix itself was investigated for die gas-phase O2 oxidation of p-xylene. The borosilicate molecular sieve, HAMS-1B-3, was partially deboronated to prepare two DBH samples containing 1.0 wt% Β (sample DBH I) and 52 ppm Β (sample DBH II). These samples were subjected to the p-xylene oxidation under the comparable reaction conditions. The results are compared with those obtained over CVD Fe/Mo/DBH (see in Table V.) The catalytic activity of the DBH sample depends on the acidity of the sample. It is in turn closely related to the boron content removed from theframeworkof the DBH. The byproducts such as toluene and pseudocumene were formed via disproportionation, and trimethylbiphenylmethane was produced via oxidative head-to-tail coupling of p-xylene. The acid sites of the catalyst are responsible for these side reactions. However, it was surprising to observe that only p-tolualdehyde without accompanying terephthaldehyde was formed over these DBH samples, and that the oxidation was also promoted by die co-presence of C 0 in the 0 feed as well as under the anaerobic condition in the flow of C 0 alone. Since the specific acid sites on DBH which are responsible for the side reactions are assumed to be occupied by the catalytic guest moieties, 2

4

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

124 Fe2(Mo0 )3 and M0O3, during the catalyst preparation step, it is logical to believe that both "single" and "paif sites are provided by the catalyst moiety rather than die DBH matrix itself with the CVD Fe/Mo/DBH catalyst. 4

The peroxocarbonate intermediate The cyclic peroxocarbonate complex is proposed as an active species for the promoting function of C0 for the gas-phase Q2 oxidation of alkylaromatics based on the above results observed with CVD Fe/Mo/DBH(1,43,47). The nature of die peroxocarbonate anion (COO bas been disclosed for the liquidphase oxidation of styrene and tetrahydrofuran by Aresta et al. (11-13) and for active metal oxide materials for selective oxidative coupling of methane by Dubois et al., (46,48,49).

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

2

Table V· Oxidation of p-Xylene over Deboronated HAMS-1B-3 (DBH) Host Matrix Catalyst Boron wt% Temperature °C Feed type

DBH I 1.0 350 3 2

DBH II 52 ppm 350 1 2

Conversion %

3.8

3.4 20.1

18.5

Product distribution (mole%) Toluene 67.6 48.4 51.1 48.7 Pseudocumene 11.6 4.5 12.7 8.7 Benzaldehyde 0 0 0 0 p-Tolualdehyde 13.2 28.2 10.3 20.9 Terephthaldehyde 0 0 0 0 Trimethylbiphenylmethane 7.6 5.0 25.9 21.7 CO 0 14.0 0 0

-

CO2

-

-

-

CVD Fe/Mo/DB 540 ppm 350 1 2 41.2

65.6

3.5 0 2.4 50.2 23.5 1.7 0.6 15.6

4.8 0 2.7 40.9 33.6 0.6 3.6

-

feed 1:0.1% p-xylene, 1.0% N ,1.0% 0 in C0 ; feed 2:1% p-xylene, 1.0% N 1.0% 0 in He; feed 3:0.1% p-xylene, 1.0% N , in C0 anaerobic condition. 2

2

2

2

2

2

2

As shown in scheme 1, it is thought that the metal-Cfe complex is initially

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

125 formed on the Fe (Mo0 )3 moiety (most likely on the four coordinated Mo center) when the CVD Fe/Mo/DBH catalyst is exposed to the feed stream containing both O2 and CO2 A CO2 molecule is then inserted into the O O bond on the resulting metal-0 complex to form an active peroxocarbonate species. It may play a key role in boosting the oxidation reaction as a one-oxygen transfer agent mimicking the mono-oxygenase enzyme. It has also been reported that the oxygen atoms directly bonded to the metal transfer to the oxophile (11) to form aldehyde and other oxygenates. 2

4

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

2

A

Cyclic peroxocarbonate Scheme 1 Conclusions A synergistic cooperation between the guest catalytic moiety, Fe (Mo0 )3 with the host DBH matrix in CVD Fe/Mo/DBH gave rise to a novel catalytic activity for the gas-phase 0 oxidation of alkylaromataics. The catalyst exhibited a unique property' to activate C O 2 to function as a co-oxidant The peroxocarbonate complex intermediate has been proposed to be an active 2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

4

?

126 species for the initial abstraction of H from the alkyl side Chain substitueras of alkylaromatics. In a typical gas-phase O2 oxidation run of p-xylene, the catalytic activity was remarkably enhanced in the co-presence of C 0 in the 0 feed gas stream. The CO2 molecule functions not only to enchance the oxidation rate, but also to improve the selectivity to the aldehyde products by significantiy lowering the reaction temperatures and suppressing die burning. It also plays a role as a reaction modulating agent to alter the product distributioa Two active centers named "paii" site and "single" site have been postulated for the formation of terephthaldehyde and p-tolualdehyde respectively from the p-xylene. These sites were preferentially poisoned by the chemical modification techniques such as Ag-doping for the "pair" site and CVD silica deposition for the "single" site. The CQ2 promoting function remained intact with Ag/Fe/DBH while it disappeared almost completely over Si/Fe/Mo/DBH. For the oxidation of ethylbenzene, the promoting function of C 0 was effective for the initial abstraction of hydrogen causing dehydrogenation to produce styrene. Contrary to the homogenous catalysts studied previously for oxidation of styrene and tetrahydrofuran, carbon dioxide remained as an inert diluent over the CVD Fe/Mo/DBH catalyst. In short, the promoting function of CC^ ties directly to the initial abstraction of hydrogen atom(s) from alkyl side chains of alkylaromatics with CVDFe/Mo/DBH.

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

2

2

2

References 1. Yoo, J.S.; Donohue, J.A. Kleefisch, M.S.; Lin, P.S.; Elfline, S.D. A p p l . C a t a l . A 1993,

105,

83

2. Aresta, M. et al. Inorg. Chem. 35(4), 1996, 4254. 3. Kolomnikov, I.S.; Lysyak, T.V. Russian Chem. Rev. 1990, 59(4), 344-360. 4. Wright, C.A.; Mathew, T.; McGill, J.W.; Gong, J.K. Fourth Intern. Conf. on Carbon Dioxide Utilization, Sept. 4-11, 1997, Kyoto, Japan, p-046, 5. Ishitani, O. et al. J. Chem. Soc., Chem. Commun. 1994, 367. 6. Keene , F.R.et al. Coord. Chem. Rev. 1985, 64, 247. 7. Jezowska-Trzebiatowska et al. Organomet. Chem. 1974, 80, C27. 8. Galan. G. et al. J. Phys. Chem. 1997, 101, 2626. 9.

Mirzabekova. S.R. et al. Kinet. and Catal. 1995, 36(1), 111.

10. Aida, T. Shokubai. 1991, 33(3), 230. 11. Aresta, M.; Fragale,C.;Quaaranta, E.; Tommasi,I.J. Chem. Soc., Chem. Commun.,1992, 315. 12. Albano, P.; Aresta, M.; Manassero, M. Inorg. Chem. 1980, 19, 1069 13. Aresta, M. et al. Inorg. Chem. 1996, 35, 4254. 14. Nishiyama, Τ and Aika, K. J. Catal. 1990, 122, 346.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

127

15. Tanaka,I..et al., Catal. Today, 1998, 45 (1-4), 55. 16. Nakagawa, K.; Kajita, C.; Okamura, M.; Kato, S.; Kasyua, H.; Ikenaga, N.;Kobayashi, T.; Suzuki, T.Catal.Lett.2000, 64, 214. 17. Mimura, N.; Saito, M. Catal. Today 2000, 55, 173. 18. Minura N.; Takahara, I.; Saito, M.; Hattori, T.; Ohkuma, K.; Ando, M. Fourth Intern. Conf. on CO Utilization, Kyoto, Japan, Sept. 4-11, 1997. P­ -024. 19. Chang, J-S.; Park, S.-E.; Park, M.S.; Anpo,M.;Yamaashita, H. Fourth Intern. Conf. on CO Utilization, Kyoto, Japan, Sept. 4-11, 1997, P­ -013. 20. Hattori, T.; Yamauchi, S.; Komatsuki, M. Proc. of the Intern. Symp. on Chemical Fixation of Carbon Dioxide, Dec. 2-4. 1991, Nagoya, Japan, Ρ 28;p429, 21. Yamauchi, S. et al. Sekiyu Gakkaishi, 1994, 37, 278. 22. Park, S-E et al. Am. Chem Soc. Meeting, Preprint, Fuel Div. Paper C45, May 1997. 23. Hattori, T.; Yamaguchi, S.; Satsuma, Α.; Murakami,Y.Chem. Lett., 1992, 629. 24. Llorca, J.; de la Piscina, R.; Sales, J.; Homs, N. Fourth Intern.Conf.on Carbon Dioxide Utilization, Sept. 7-11, 1997, Kyoto, Japan, O-03. 25. Hackerman, N. et al. J. Electrochem. Soc. 1984, 131, 1511. 26. Yoshida, Y. and Ishii, S. Proc. of the Intern. Symp. on Chemical Fixation o Carbon Dioxide, Dec. 2-4. 1991, Nagoya, Japan,p29;p423. 27. Aresta, M. J.Mol.Chem. 1987. 41. 355. 28. Kusama, H. et al. Catal. Today, 1996, 28, 262. 29. Ihaara, T. et al. Bull. Chem. Soc. Jpn. 1996, 69, 241. 30. Tominaga, K et al. Bull. Chem. Soc. Jpn. 1995, 68, 2837. 31. Inui et al., Appl.Catal.1982, 2, 389. 32. Aresta, M. Quaramta. E., Chem. Tech. 1997, 27(3), 32. 33. Toda, T,; Yoshida, M.; Ohshima, M. Proc. of the Intern. Symp. on Chemical Fixation of Carbon Dioxide; Dec. 2-4. 1991, Nagoya,Jpn.B8,p185. 34. Aresta, M, Intern. Symp. on CO Conversion and Utilization in Refinery and Chemical Processing,219 National meeting, ACS, Sa Francisco, CA, March 26-31, 2000, Preprint, Div. of Petrol. Chem., 45(no.1), 92. 35. Centi, G.; Misono, M., Catal. Today, 1998, 41, 287. 36. Yoo, J.S.; Donohue, D.A.;Choi-Feng, C. Advanced Catalysts and NanostructuredMaterials;Moser, W.R., Ed., Academic Press: New York, NY, 1996; Chapter 17, pp 453-477. 37. Zajac, G.W.; Choii-Feng,C.;Faber, J.; Yoo, J.S.; Patel. R.; Hochst, H. J. Catal. A 1995, 151, 338. 38. Yoo, J.S.; Lin, P.S.; Elfline, S.D.Appl.Catal. A, 1993, 106, 259. 39. Yoo, J.S.; Donohue, J.A; Kleefisch, M.K.Appl.Catal. A, 1994, 110, 75. 40. Yoo, J.S.; Choi-Feng,C.;Donohue,J.A.Appl.Catal. A, 1994, 118, 87. 2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

2

2

th

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

128

41. Yoo, J.S.; Lin, P.S.; Elflline, S.D. Appl. Catal. A, 1995, 124, 139.

42. Yoo, J. S. Appl. Catal. A, 1996. 135, 261. 43. Yoo, J. S. Appl. Catal. A, 1996, 143, 20. 44. Yoo, J.S. Appl. Catal. A, 1996, 142, 19. 45. Yoo, J.S.; Choi-Feng, C.; Zajac, G.W. Appl. Catal. A, 1999, 184, 11. 46. Hayward, P.J. et al. J. Am. Chem. Soc.1970,92(20), 5873.

47. Yoo,J.S.Appl. Catal. 1993, 106, 259 48. Dubois, J.-L.; Bisiaux, M. Minoun. H.; Cameron, C.J. Chem. Lett. 1990, 6, 967. Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch008

th

49. Dubois, J.-L.; Cameron, C.J. Proceedings of the10 Intern. Congr. On Catal. New Frontiers in Catal. Guczi, L.; Solymosi, F; Tetenyi, P.; Ed.; Part-C, Akademiai Kiado, Budapest 1992, pp 2245-2252.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 9

Effective Conversion of CO to Valuable Compounds by Using Multifunctional Catalysts Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch009

2

Tomoyuki Inui Gas and Chemical Research Division, Air Water Inc., Sakai 592-8331, Japan (Fax: (+81)-722-44-8085; email: [email protected])

Indispensable conditions for C O mitigation by catalytic conversion are enumerated. These are very rapid conversion rate and high selectivity to valuable compounds. Since reduction of C O needs expensive hydrogen, the ways to get hydrogen inexpensively are described. One is the reduction of C O by methane or natural gas instead of hydrogen. In order to realize this reaction without coke formation, the four-component composite catalyst, the Rh­ -modified Ni-based catalyst with Ce O and Pt as additives has been developed by author et al. Another is the on-site heat supply by catalytic combustion to compensate the large endothermic heat of reforming of methane. Especially, the addition of more easily combustible hydrocarbons such as ethane, propane, and butane, which are contained in natural gas, made possible marked decrease in furnace temperature around as low as 570 - 600 K . Even such lower furnace temperatures, the catalyst temperature rises up to around 970 Κ and an equilibrium conversion of methane is observed even at a very short contact time such as 5 msec. Simultaneous reforming of methane with CO and H O also achieved. Ultra rapid methanation of C O on the Ni-based three­ -component composite catalyst, Ni-La O -Ru, highly effective 2

2

2

2

3

2

2

2

2

130

3

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

131 synthesis of methanol by C O hydrogenation on the Pd-Ga-modified Cu-Zn-Cr-Al-Ox-Al-Ox catalyst, the effective synthesis of ethanol by C O hydrogenation on the multifunctional catalyst are then summarized. Finally, highly effective syntheses of light olefins and gasoline via methanol synthesis from C O hydrogenation using multi-step reactors connected in series are described. The sequential results of these catalytic conversions of C O shows a high potential to realize the processes for mitigation or recyclic use of C O . 2

2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch009

2

2

Hence carbon dioxide is the final combustion product of organic compounds, C 0 itself has little value, and in order to obtain new products by chemical reduction, the large amount of additional energy, in particular expensive hydrogen, is necessary. However, the catalytic hydrogénation of C 0 is superior to other chemical conversion methods for C 0 . Because, C 0 can be converted with an extremely higher rate on the well designed solid catalysts than other chemical conversion methods, and desired compounds can be synthesized very selectively. The largest problem in the catalytic hydrogénation would be concentrated in the effective and economic production method of the huge amount of H as the reducing reagent for C 0 . Development of the highly active Ni-based reforming catalyst which works at much lower temperatures than that of the conventional catalyst will be introduced first. Since the novel catalyst has high activity for not only steam-reforming but also C0 -reforming of methane, H for C 0 hydrogénation could be replaced by C H and H 0 . In this review paper, the focus will be concentrated to the C 0 reduction by H or C H to synthesize highly valuable major building blocks for petrochemical industries such as ethylene, propylene, methanol, ethanol, and high quality gaseous and liquid fuels such as substituted natural gas and high octane-number gasoline. These high value products have a potential to partly compensate the cost of reducing reagents, hydrogen or methane, and these C0 -conversion routes lead to a new paradigm for chemical industries and energy usage cycles in many fields. 2

2

2

2

2

2

2

2

4

2

2

2

2

4

2

Rapid C0 -Reforming of Methane 2

The major conventional production method of H is the steam reforming of saturated hydrocarbons, in particular natural gas or methane, on the stabilized supported N i catalyst, and the reaction is operated at a high temperature around 1,170 K . In order to moderate the coke deposit, an excess concentration of steam than the stoichiometry of the reaction, Eq. [1] is added in the feed. A s the conventional catalyst has no ability for C 0 2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

132

activation, C 0 once formed by the shift reaction, Eq. [2] via Eq. [3] cannot be converted to other molecules by the reaction between C 0 formed and methane unconverted as Eq. [4]. 2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch009

CH CO CH CH

4

2

2

2

4

4

165 kJ/mol -41 kJ/mol ΔΗ° 206 kJ/mol Δ Η ° = 248 kJ/mol

+ 2H 0 4H +C0 , + H 0 -* H + C0 , + H 0 - * 3 H + CO, +C0 2 H + 2CO, 2

2

ΔΗ° ΔΗ°

2

2

2

2

2

[1] [2] [3] [4]

In recent years however, C0 -reforming of methane as expressed by E q . [4] has gathered a great deal of attention in the sequential international congresses on C 0 mitigation and/or utilization (1-10). Reflecting these situations, several elaborative reviews on this subject have been made by Fox III (11) in 1993, Rostrup-Nielsen (12) in 1994, Ross et al. (13) in 1996, Halmann and Steinberg (14) in 1999, and Bradford and Vannice (15) in 1999. Especially, the last one (15) summarized as many as 190 papers including classic papers. The most recent 16 papers including in this review were published in 1997. The focus of this review is concentrated mainly on the fundamental aspects such as activation of C H and C 0 , carbon deposition, kinetics, and reaction mechanisms. However, catalytic reaction-engineering aspects involving autothermal reforming and co-reforming such as C 0 + 0 and C 0 - H 0 - 0 are not mentioned. From the view points of the rapid reforming, which is applicable to industrialization, these reactionengineering aspects are indispensable, and indeed, in the most recent few years, the papers treated on these subjects are rapidly increasing (16-21). Including the most recent papers published until middle of 2000, the review article written by the author (22) on "Reforming of C H by C 0 Associated with 0 and /or H 0 " will be published from the Royal Society as Catalysis Vol. 16 in 2001. Therefore, here, only the essence of our recent research works on C0 -reforming is described below. It seems to be contradiction that in order to obtain H for C 0 hydrogénation, the additional fossil fuel is necessary to maintain the high reaction temperature, and moreover C 0 is produced as the by-product. However, if by the innovative improvement in catalyst structure, on which no coke deposit occurs and exhibits a very high reaction rate even at the low temperature range around 600 K , situation would become different. Because the heat to maintain such a medium-range temperature could be supplied by the waste heat of large scale facilities of industries and even by the accumulated heat of solar energy. Furthermore, the large endothermic heat of reaction could be supplied on-site by the catalytic combustion of more easily combustible hydrocarbons and/or a part of methane fed on the same catalyst surface, on which methane reforming reactions are advancing. In order to realize the rapid synthesis of hydrogen through methane reforming, the synergistic effect of composite catalysts and the combined reactions to compensate the large amount of endothermic heat were investigated. A Ni-based three-component catalyst such as Ni-Ce 0 -Pt 2

2

4

2

2

2

2

2

2

4

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

3

133 supported on alumina-wash-coated ceramic fiber in a plate shape (23) was very suitable for the reforming of methane. The catalyst composition was set at 10 wt% N i , 5.6 wt% L a 0 , and 0.57 wt% Pt for example, or molar ratios of these components were 1: 0.2: 0.03. Even with such a low concentration, the precious metal enhanced the reaction rate markedly, and this synergistic effect was ascribed to the hydrogen spillover effect through the part of precious metal and it resulted in a more reduced surface of the main catalyst component. In particular, a marked enhancement in the reaction rate of C0 -reforming of methane was observed by the modification of a low concentration Rh to the Ni-Ce 0 -Pt catalyst (24). Fig. 1 shows that the four-component composite catalyst ( © in Fig. 1) exhibited a very high activity for C0 -reforming of methane, Eq. [4], with the stoichiometric ratio of the products approaching to the reaction equilibrium even such as a high space velocity (SV) 730,000 h" , or a very short contact time of the reaction gas of 5 msec, on the basis of the net volume of the catalyst and its support. The activity of the composite catalyst was much larger than the sum of the activities of the component catalysts, namely the Rh and the Ni-Ce 0 -Pt catalyst. This means that the composite catalyst involved a two-step spillover, i.e., hydrogen formed was adsorbed on the Rh part very rapidly, faster than on the Pt part, and then the spiltover hydrogen was abstracted by the Pt part, this being followed by its diffusion toward the major catalyst component, the N i part. Consequently, the N i part can be kept in a reduced surface state and the rapid reaction progress can occur on it. The role of C e 0 would be not only to promote dispersion of the N i component, but also to act as a transporting media for spiltover hydrogen. Hydrogen and C O were obtained in equivalent amounts by the reaction, and the space-time yield of H and C O obtained at 873 Κ was 7,690 mol/1 · h at 59% methane conversion. In a reaction of C H - C 0 - H 0 - 0 on the four components catalyst, an extraordinarily high space-time yield of hydrogen, 12,190 mol/l-h, could be realized under the conditions of very high space velocity (5 msec) (25). In order to overcome the restriction in conversion of the reaction equilibrium, the combustion of more easily combustible hydrocarbons such as ethane or propane, which are involved in the natural gas, was then combined by adding these hydrocarbons and oxygen. The aimed reactions to produce only syngas are expressed as the following Equations [5] and [6], including generation and consumption of heat by combustion and reforming, respectively. 2

3

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch009

2

2

3

2

1

2

2

3

3

2

4

2

2

2

(x+5y) C H + χ C 0 + y C H + 3.5y 0 4

2

2

6

ΔΗ (x+7z) C H + χ C 0 + ζ C H + 5z 0 4

2

3

8

2

ΔΗ

-> (2x+7y) C O + (2x+13y) H

2

7 7 3 Κ

= 259x - 240y kJ/mol

-> 7 7 3 Κ

[5]

(2x+10z) C O + (2x+18z) H

= 259x - 375y kJ/mol

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

2

[6]

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch009

134

Φ SV 73,000b- (CT 49.3ms) 1

10

20

30

40

C H conversion

(%)

4

Figure 1. Synergistic effect in Rh-modified Ni-Ce 0 -Pt catalyst caused by combination with each ingredient upon the C0 ~reforming of methane Catalyst supported on alumina washcoated ceramic fiber Φ 10.0wt% Ni - 6.0wt% Ce 0 Φ 1.0wt% Ft, Φ 0.2wt% Rh, φ lO.Owtfo Ni - 6.0wt% Ce 0 - 1.0wt% Pt, ® 10.0wt% Ni - 6.0 wt% Ce 0 - 0.2wt% Rh, @ lO.Owtfo Ni - 6.0 wt% Ce 0 - 1.0wt% Pt - 0.2wt% Rh Feed gas: 10 mol% CH -10 mol% C0 - 80mol% N 2

3

2

2

3f

2

2

2

4

SV:WBÊ 73,000 K ; È23 1

3

3

3

2

2

1

730,000h

Dotted line: Equilibrium conversion level at 873 Κ

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

135

φ

^

(66.2%), 2,290 mol/l-h (36.2%), 16,800 mol/l-h

φ

Mffi

(82.2%), 3,400 mol/l-h (47.8%), 20,360 mol/l-h

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch009

mm

(80.8%), 3,160 mol/l-h (57.3%), 24,870 mol/l-h

0

5,000

10,000

JL

15,000

J_

J-

20,000

25,000

Space-time yield of syngas (mol/l-h) Figure 2. Effect of catalytic oxidation of ethane or propane on the C0 reforming of methane at two kinds of space velocities 73,000 and 730,000 h Catalyst: The Rh-modified four-component catalyst Gas composition: φ 35% CH -10% CO - 55%N φ 35% CH -10% C0 - 5% Cflt -17.5% 0 - 32.5% N φ 35% CH -10% C0 - 3.3% CsH -16.5% 0 - 35.2%N Catalyst-bed temperature: 700 °C (Furnace temperature: 500 °C) Pressure: 1 atm. Numerals in parenthesis are methane conversion. 2

1

4

2

4

2

4

2

2

2

8

2

2

2

In case of ethane or propane addition in C0 -reforming of methane, the catalyst temperature abruptly rose even at a very lower furnace temperature such as below 625 K, and it was maintained at much higher temperature than the furnace temperature, indicating that by the in-situ heat supply due to the catalytic combustion, the methane conversion was eventually induced at a very lower temperature range. As shown in Fig. 2, an extraordinarily high space-time yield of syngas, as high as 25,000 mol/l-h was obtained at a considerably low furnace temperature around 773 Κ (26). 2

Rapid C0 -Methanation 2

Catalytic hydrogénation of carbon oxides was found by Sabatier and Senderens (27) at the beginning of twentieth century. Since then, a numerous studies have been done (28, 29). Industrially, this reaction has been applied to the purification of H -rich gas and production of substituted natural gas (30). Two decades ago, methanation reactions, i.e., inverse reactions of Eqs. [3] and [1] were applied to energy saving devices and processes. A typical one known as EVA-ADAM process (30) was achieved in Germany. The 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

136 process consists of two reaction systems. The one (EVA) is endothermic steam reforming of methane, i.e., Eqs. [1-3]. this requires a large amount of energy at a temperature level of more than 1,200K, which is supplied form the high temperature gas furnaces. Another one (ADAM) is exothermic methanation of CO and C 0 , i.e., inverse reaction of Eqs. [3] and [1], where heat is set free over a wide range of temperature. The energy taken up in EVA is transformed into a cold H -rich gas that is transported to ADAM, where heat is released according to the requirements of the heat market. Recently, C 0 methanation has had a new significant object concerning with the global warming problem. Since methanation of C 0 needs four times molecules of H , it seems to be disadvantageous. However, as the rate of C 0 methanation is far beyond all other methods for chemical conversion of C 0 to hydrocarbons and oxygenates. The largest consumption of hydrogen per unit C 0 mole equivalents to that the largest energy taking into methane molecules. Therefore, the methane synthesized by hydrogénation of C 0 is expected as the transportation media of the energy injected to the reaction. When the rapid C 0 methanation progress at a low temperature range around 470 - 570 K, the large exothermic reaction heat itself is available, and furthermore, when the methane formed is used as the fuel, a high temperature even 1,300 Κ can be obtained. Therefore, it is thought that the rapid C0 -methanation is significant as one kind of the chemical heat pump, by which a low temperature, i.e., a low-value energy, can be transformed into a high temperature, i.e., a high-value energy, through chemical reactions. Most recently, Halmann and Steinberg reviewed C 0 methanation (32) in their monograph. As shown in this review, the most of the studies on C 0 methanation used supported single component catalysts, except small number of studies, such as Co/Cu/K catalyst (33), Rh/Ce0 /Si0 (34), FeMn doped with Rh and La (35), Fe-Zr-Ru (36), and Ni-La 0 -Ru supported on spherical silica or alumina reported by the author et al. (37). The last one had been developed more than two decades ago, however, this catalyst is regarded as one of the most active methanation catalysts among the whole catalysts, as recognized by Halmann and Steinberg in their review. The author has described in detail the sequential results on this catalytic performance and the reason of exerting the ultra-rapid reaction rate in the review article of "Methanation" in "Encyclopedia of Catalysis" (38), which is now under preparation with editing by Professor Horvâth et al., and will be published soon. Therefore, here, only the essence of these works is described below. The author et al. have studied on the C 0 methanation since 1970's, and developed the Ni-based three components composite catalyst supported on the spherical silica support having meso-macro bimodal pore structure (37). The catalyst contained La 0 by 1/5 mol of Ni, and Ru by 1/30 mol of Ni. As shown in Fig. 3, combination of these three kinds of catalyst components exerted an evident synergistic effect on the methanation rates of C 0 , and showed the high conversion rate to synthesize methane exclusively (24, 37). 2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch009

2

2

2

2

2

2

2

2

2

2

2

2

2

2

3

2

2

3

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

137 The catalyst made possible the C O - C 0 co-methanation (39), and therefore, the process of C 0 elimination in the conventional S N G process could be omitted. The cause of the high performance was elucidated as follows; the adsorption capacity of catalyst for C 0 was increased by the weak basicity of L a 0 , and hydrogen adsorption was markedly enhanced by combining Ru, which worked as the porthole of hydrogen spillover. The meso-macro bimodal pore structure has the roles of the supported bed for catalytic substances and the pass for quick diffusion of the reactants, respectively (37). 2

2

2

Downloaded by PENNSYLVANIA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch009

2

3

5.0wt%Ni | 0.7wt%Ru

l

a

Ratio ofSTY >

11.4

5.5wt%-0.8wt%Ru 4.6wt%-2.6wt%La 0 2

3

4.3wt%Ni-2.5wt%La O -0.7wt%Ru 2

3

0

10

Relative methanation activity Figure 3.

Activity of C0 methanation for the Ni-La CH OH+ H 0

(1)

CO + H 0 - > C 0 + H

(2)

2

3

2

2

2

2

2

CO + 2 H -> CH OH 2

(3)

3

As shown in Figure 2, activity ofCu-Mn-0 for methanol synthesis from CO/H is lower than that in Figure 1 from CO/C0 /H . C 0 enhances the activity of the oxide. 2

2

2

2

100

0 Figure 2. STY from CO/H

20 40 60 80 C u / (Cu + Mn) (mol%) 2

100

as a function of Cu Content of Cu-Mn

Oxide.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

158 On the other hand, Cu-Mn oxide was mixed with γ-alumina (1:1 by weight) and dimethyl ether was formed instead of methanol by the reaction (4). 2CH OH 3

(4)

2

As shown in Figure 2, the activity was increased to the same level of methanol formation from C O / C 0 / H . The result suggests that water produced as by­ product of reaction (4) is transformed quickly by water-gas shift reaction (eq.(2)) on the oxide and the produced C 0 enhances the activity of Cu-Mn oxide. Like onCu/Zn/Al catalyst, both reaction (1) and (2) proceeds on Cu-Mn oxide. Furthermore, activity of Cu-Mn oxide for methanol synthesis is proportional t o C u area as shown in Figure 1, i.e., Cu° is active site of Cu-Mn oxide. It was reported that the active site for C 0 hydrogénation is Cu° and that bothCu andCu are necessary for C O hydrogénation (11). From these points of view, it can be concluded that hydrogénation of C 0 is predominant on CuMn oxide catalyst for methanol synthesis from syngas containing C O . 2

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch010

CH3OCH3 + H 0

2

2

0

2

0

+

2

2

Catalysis of Supported C u - M n Oxide The activity of bulk catalyst prepared from oxalate salts was over 90gM e O H /kg-cat/has shown in Figure 3. The advantage of oxalate-ethanol method is clearly shown over nitrate method. The other activity than that of bulk oxide was calculated based on the loaded oxide weight. From this result zirconia and titania are good candidates. But unfortunately the surface area are not so high, so, the activity based on total weight is not good. High surface area for zirconia or titania is required.

Preparation

of ZrU2 and T1O2 with High Surface Area

Z r 0 with high surface area was prepared from zirconium hydroxide with the addition of small amount of phosphoric acid (7). In the case of titania, similar technique was available. After titanium hydroxide was first prepared from titanium alkoxide, the conditions for the hydrolysis was also optimized. The standard procedure is as follows: water was added at 55°C into 10mol% titanium butoxide in i-propanol to hydrolyze the butoxide. After centrifuge the precursor was calcined at 450 °C. As a result TiO 2(111) was prepared. As shown in Figure 4, while solvent forthe butoxide, concentration of the butoxide and temperature for hydrolysis hardly influence on the surface of the 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch010

159

Figure 3 Activity of Cu-Mn-oxide on commercial support.

Figur e 4. Pr epar at ion of Ti Ο suppor t. 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

160 resulting oxide, addition of Si or Ρ gives good result. TiCh(219) were prepared, respectively.

Thus TiO (206) and 2

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch010

Catalysts of Cu-Mn-0 / Zr02 In contradistinction to the activity of the bulk catalyst normalized by weight of Cu-Mn oxide, T1O2 and Z1O2 supported catalysts show good results as in Figure 3. Hereafter, the catalytic feature of Ζ1Ό2 supported one was investigated. Figure 5 shows the effect of oxide content onZr0 (214) andZrCh (81). The activity of the catalysts is almost same and increases in proportion to the oxide content under 30wt% which corresponds to the monolayer content of spinel on Zr02(81) support. Above 30wt% the activity of Cu-Mn oxide on Ζ1Ό2 (214) increases on a proportional basis while that on Zr02(81) levels off. The result suggests that spinel oxide spreads in single-layer on theZr02 surface until the whole surface is covered. 2

From X R D of the supported catalyst (Figure 6) no spinel oxide is found on zirconia and only copper oxide was observed. The result suggests that copper was sintered at high loading while manganese oxide was amorphous. Effect of Cu/Mn ratio was investigated on Zr0 (214) which contains the same molar amount of Cuand/or Mn oxide with 50wt% Cu/Mn=l/1 oxide. Similar synergy is observed as shown in Figure 7 on the supported Cu-Mn oxide as the bulk CuM n oxide and the activity shows maximum at Cu/Mn=l/1. The result strongly suggests that synergy between Cu andMn appears on ZrO 2(214). 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

161

Cu-Mn-0 loading 70%

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch010

A•

•

50% '^ ^^Λ^^ ΜΜ»^Λ«^^ ν

30

28%

|

35

2Θ

40

45

Figure 6. XRD of Cu-Mn-0 on Zr0 (214). 2

0

20 40 60 80 Cu/(Cu+Mn) (mol%)

100

Figure 7. Cu-Mn synergy on Zr0 (214). 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

162 Catalysts of Cu-Mn-01 T1O2 On titaniasynergy of copper and manganese was also observed. But the highest activity was achieved at Cu/Mn ratio 2 to 1 as shown in Figure 8.

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch010

60 ,

0

,

20 40 60 80 Cu/(Cu+Mn) (mol%)

100

Figure 8. Cu-Mn synergy on Ti0 (219). 2

In the case of titania, X R D gave much information as shown in Figure 9. Titania is anatase type and all the particle size o f T i 0 was same after oxides were supported as shown above in Figure 10. The particle size was also same as calculated from B E T surface area (72Â). This result suggests that most of titania is in anatase form without amorphous one. Other species was CuO. Intensity of CuO is low at low loading and increases above the monolayer loading rapidly while B E T surface area decreases as shown below in Figure 10. Large particle of copper oxide may retard the reaction and decrease the activity. The behavior of the activity of Cu-Mn-O/Ti0 (219) was almost same as on zirconia as shown in Figure 11. The activity increased with the amount of loading and it was highest at monolayer loading. From observation on X R D we can imagine what happened on these supports. When the activity as a function of oxide content is overlapped we can see that below monolayer loading the activity was almost same. But above it, the activity on zirconia decreases rapidly with increasing the oxide loading. Average particle size from B E T surface of titania (72 Â) is almost two times larger than that of zirconia(48 Â ) . S ο plugging by CuO was serious problem for zirconia. Maybe this is because the low activity of zirconia at around these loading. 2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch010

163

90%

70% 60% mm*0Ι^^ΙΙΜ^^^Α»' ^

40%

0% T1O2 (Anatase)

Oil 20

30

40 Ζ θ / degree

Figure 9 XRD of Cu-Mn-0

50

60

onTi0 (219). 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.




CO

->

HCOOH

->

ROH

RH

(8)

We considered the following catalyst systems as the basis for our catalyst design. •

The homogeneous hydrogénation of C 0 to formic acid can be catalyzed by [(dppp)Rh(hfacac)] (dppp is a bisphosphine ligand and hfacac is hexafluoroacetylacetonate) in dimethylsulfoxide solvent in the presence of triethylamine at 23°C and 40 atm pressure of C 0 / H (1/1). The ratedetermining step in the catalytic cycle appears to be the liberation of formic acid (12). One of the most active homogeneous catalyst systems for formic acid synthesis comprises of Ru-phosphine complexes in supercritical C 0 solvent. A t 50°C and 205 atm, turnover frequency (TOF) of 1400 h" has been reported (13). For comparison with the same catalyst in THF, the TOF is 80 h" . The Ru (CO)i /KI homogeneous catalyst system in N-methyl-2-pyrrolidone (NMP) solvent yields a mixture of CO, C H O H , and C H at 240°C and under 8 M P a of C 0 / H (1/3) (14). Electrochemical reduction of C 0 to C O catalyzed by Pd complexes has been reported (15). Photochemical reduction of C 0 to formate catalyzed by metal complexes has been reported (16). The process uses a R u complex to capture sunlight and affect this transformation via electron relay, but the drawback is the consumption of expensive electron donor additives that are added to make the overall process catalytic. 2

2

•

2

2

1

1

•

3

2

3

2

• •

4

2

2

2

2+

The representative catalyst systems and the corresponding product selectivity are shown in Table II. It is clear that poor selectivity, low productivity, severe reaction conditions or other system limitations make these catalysts commercially unattractive. We specifically focused on developing catalysts that allow homogeneous liquid phase synthesis of methanol from C 0 : 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

3

Slurry

Ru(acac)

Homogeneous

Highly dispersed slurry

Fe/KOR 9 25

34

60

66

66

40

40

50

60

50

-

MeOH/ Glyme E-164/ Peg-400

THF

3

DMSO/ Et N THF

Oil

Solvent

3 0

250

120

270

270

23

260

Τ °C

3

2

20

130

4

5

Ρ MPa

r

3

3

3

4

CH OH >99% HCOOH > 99% CH OH > 80% HC > 90% CH OH > 99% C C -OH > 90%

Product

1

25

17

2

2

13

6

Reference

*0.5L A E Zipplerclave Reactor dppp = bis(diphenylphosphino)propane; hfacac = hexafluoroacetylacetonate; H C = hydrocarbons; L = Ligand; K O R = potassium alkoxide; E-flo-164 = C hydrocarbon solvent; Peg: polyethyleneglycol.

x

50

50

2

2

Feedgas . %CO %H

%co

NiL /KOR

3

Homogeneous

Homogeneous

Slurry

Lurgi

(dppp)Rh(hfacac) Ru(acac)

Mode

Catalyst

Table II. Liquid Phase Catalytic Hydrogénation of CO2/CO

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch011

175 C0

2

+ 3H

CH3OH

->

2

+

1

H 0

A H ° 2 9 8 = - 52.8 kJ.mor

2

(9)

Our approach to C 0 hydrogénation is based on the liquid phase low temperature (LPLT) concept that was specifically developed at BNL to design enhanced carbon utilization catalyst for syngas conversion into methanol (77). 2

CO

+

2H

2

->

CH3OH

ΔΗ°

2 9 8

= - 128.6 kJ.mol

1

(10)

With C 0 as the reactant, the challenge was to initiate the water-gas-shift (WGS) reaction that can be combined with Reaction (10) to yield methanol. Since carbonyls of Mo, Co, Cu, Pt, Ru are known WGS catalysts, we conducted a preliminary screening study with these complexes to assess their ability to reduce CO to methanol. Peg-400/base or TEA itself provided the basic medium in these runs. The data are shown in Table III. Runs conducted in the batch unit with initial syngas (CO/H ) charge of 5 MPa showed that except for Co, all other metal carbonyls evaluated consumed syngas. The gas consumption rate was 2 mmol/min at 130°C with Cu and 0.7 mmol/min at 150°C with Mo. The gas chromatographic analysis indicated that methanol was indeed formed in all these runs. Further work is underway to select the best metal that will yield high methanol selectivity under mild conditions of temperature (< 150°C) and pressure (< 5 MPa).

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch011

2

2

Table III. Catalytic hydrogénation screening runs Catalyst

T,

Cu(OMe) K PtCl Co (CO) Ru (CO) Mo(CO) 2

2

6

2

8

3

12

6

°c

130 150 130 160 150

1

Rate, mmol/min

Gas Consumed ,mmol

1.2 2.0 0 1.2 0.7

230 190 —

185 201

0.5 L A E Zipperclave batch unit; catalyst: 1 mmol; Base: Potassium methoxide = 100 mmol; solvent: TEA or Glycol = 120 mL; Syngas: H /CO ~ 2; Pi = 5 MPa at room temperature. 2

Geologic Formations: The Ultimate Slurry Reactor for Catalytic CO2 Reduction In the previous section, we described the results of our laboratory study to develop a highly efficient catalyst for methanol synthesis via C 0 hydrogénation 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

176 that operates in an amine or a glycol solvent. Successful development of this novel system has the potential of lowering the cost of C 0 sequestration by integrating the C 0 capture and utilization aspects but the cost of H remains an issue. Our aim is to design a highly effective catalyst for C 0 hydrogénation in the laboratory and apply these results to ultimately conduct C 0 reduction in geologic formations. Therefore, we envision geologic formations as slurry reactors to produce H from H 0 that is subsequently utilized to hydrogenate C 0 in these formations. We have based this approach on what is known about temperature and pressure conditions that exist in oil and gas reservoirs. We discuss below the basis of such an approach and the consequence of utilizing this pathway to recycle C 0 by its conversion into fuels. 2

2

2

2

2

2

2

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch011

2

2

Basis for the Proposed Approach Oil and gas reservoirs have broad geographic distribution in the U.S. The porous and permeable reservoir rocks in oil and gas reservoirs include limestone, dolomite, and sandstone. The cap rock and basement rock which have a far lower permeability than the reservoir rock act as a seal to prevent the escape of oil and gas from the reservoir rock. Typical cap rock and basement rocks are clays and shale, that is, strata in which pores are much finer than those of reservoir rocks. The impermeable cap rock usually consists of shales that are clay minerals. The type of oil and gas reservoir depends on the boundary structure between the reservoir rocks and the cap rock. Many oil and gas accumulations (reservoirs) are trapped in either anticlines or salt domes. The major metal elements found in reservoir rocks are Na, Ca and M g . There are many other metal elements that occur in low concentrations in the rocks, including transition metals such as Fe, N i , M o , Pd. In essence, any element that has been identified in seawater can be found in the reservoir rocks. Oil and gas reservoirs typically contain H 0 . The water may occur as wetting films around the sand grains as well as in some completely filled pores. The water in most oil and gas reservoirs is heavily salted and different in terms of detailed composition, but is generally neutral with a pH close to 7, as in most eastern reservoirs in the U.S. (18). 2

The temperature and pressure gradients of the earth vary regionally and vertically. The global average geothermal gradient is about 22°C/km, but ranges from as low as 10°C in old shield areas to as much as 50°C/km in active zones of sea floor spreading (19). Geothermal gradients in sedimentary basins

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

177

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch011

generally range from 15-50°C/km; the average may be taken to be 30°C/km (20). The pressure gradients range from 0.434 psi for fresh water, 0.465 psi per foot for "standard" water containing 100 parts per thousand of dissolved solids, and 0.50-0.55 psi per foot for strong brines. The geostatic pressure depends on the bulk wet densities of the rocks including their fluids. The higher end of the overburden pressure gradient is 1 psi per foot (27). The estimated geologic conditions of many oil and gas reservoirs in Pennsylvania range from 1.1 to 5.7 miles in depth, 16°C to 150°C in temperature, and 17 to 86 M P a pressure. The role of the sedimentary sulfur cycle in the diagenesis of organic matter is of wide interest. The sulfur content in crude oils varies from < 0.05 to 14 wt%. The subject has been reviewed in a recent book (22). The foregoing discussion sets the stage for the proposed concept. The essential elements of a catalytic system that can convert the sequestered C 0 into liquid fuels in geological formations are as follows: 2

•

•

•

The cap rock and basement rock, which have a far lower permeability than the reservoir rock act as a seal to prevent the escape of fluids from the reservoir rock and can be viewed as a large natural reactor vessel. The reservoir rock must possess fluid-holding capacity (porosity) and fluid transmitting capacity (permeability). Transition metals such as Fe, N i , M o , Pd, that are normally found, though at ppm or ppb concentrations, in the reservoir rocks can be viewed as highly dispersed supported heterogeneous catalysts. The presence of oil and H 0 can be viewed as the available reaction solvents. H 0 will also serve as a natural source of H . The decreased concentration of 0 as a function of depth provides a natural reducing atmosphere for the proposed consecutive reduction reactions, i.e. reduction of H 0 to H followed by C 0 / H conversion to H -rich fuels. A t the reservoir depth of interest, the temperature and pressure range from: T: 16-150°C and P: 17-86 MPa. Sulfur could be implicated as a catalyst in the formation of hydrocarbons. 2

2

•

2

2

2

• •

2

2

2

2

Some direct supporting evidence for the catalytic effects of certain metals in naturally occurring rocks and the feasibility of H production from water by the action of transition metal species can be found in several recent papers (23,24). In these studies, it was shown during a one-year experiment that carbonaceous rocks (Monterey Formation, California, Eocene) that were low maturity and organic- rich (composition: 14 wt% total organic C, 7.5 wt% S, 350 ppm N i and 560 ppm V ) produced C H from n-octadecene-1 and H at 190°C. The reaction is believed to proceed via the hydrogenolysis of the Ο unsaturated hydrocarbon to C H at a rate of ~10" g CH /g. rock/d. The addition of H 0 , up to ~ 2.5 wt%, 2

4

2

ΐ 8

7

4

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

178

appeared to increase both catalytic activity and product selectivity. It remains to be established the nature of the catalytically active species for H production from H 0 , which is a key issue in the proposed C 0 reduction scheme. 2

2

2

Effort at BNL AND PSU The proposed concept integrates C 0 sequestration with subsequent conversion in geologic formations. Catalytic activation of small gas molecules, namely C H , CO, C 0 and H that are major constituents of coal or natural gasderived synthesis gas is a subject of ongoing efforts both at B N L and P S U (25,26). The new concept in the proposed research seeks to constitute a new and catalytic path for C 0 conversion into reduced and useful products such as hydrocarbon fuels that only uses naturally available geothermal energy in oil and gas reservoirs following C 0 sequestration in such geologic formations.

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch011

2

4

2

2

2

2

H 0 2

-»

H

+

2

(0)

surroun

dings

(11)

In Reaction 1, the Ο from H 0 is stabilized by naturally occurring materials in the surroundings. The produced H then catalytically reacts with C 0 to produce reduced products (according to Reactions (1) and (2)) that can be recovered after prolonged storage in geologic formations. The initial challenge is to demonstrate the reaction sequence for C 0 to liquid fuels in micro pressure vessels under simulated geologic formation conditions in the laboratory. For this study, two Ni-rich natural rock samples, Niccolite (Ni As ) and Pentlandite (Fe, N i ) S , have been procured and their characterization is underway at the B N L National Synchrotron Light Source (BNL/NSLS). Runs are planned with these natural rocks in a micro-reactor to demonstrate: 1) catalytic production of H from H 0 and 2) catalytic hydrogénation of C 0 under the pressure and temperature conditions that are relevant to oil and gas reservoirs. The focus is on the feasibility of the production of liquid fuels from buried C 0 under natural reservoir conditions. We then plan to utilize these laboratory data and design a system to selectively enhance the desirable reactions by adding trace amounts, in a vapor or a liquid form, to the "natural reactor" i.e., a selected geologic formation site. The proposed concept of seeding the geologic formation site with a highly active catalyst to enhance the H 0 to H production and the subsequent C 0 / H reaction to produce energy liquids closes the natural carbon cycle in fossil fuels. Interestingly, the concept is an alternative to "seeding the ocean with Fe to increase the growth of phytoplankton to enhance the C 0 uptake. 2

2

2

2

x

9

2

y

8

2

2

2

2

2

2

2

2

Finally, it should be mentioned that there may be other undesirable reactions or more desirable reactions in geologic formations in addition to what

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

179 we have considered; a number of issues need to be clarified, and many fundamental questions remain to be answered by further study.

Concluding Remarks The production of H -rich synthetic fuels by catalytically recycling carbon in C 0 is an attractive C 0 sequestration option. In this paper, the C0 -capture step that typically utilizes an amine or glycol solvent is integrated with subsequent catalytic hydrogénation of C 0 to methanol. The challenge is to design a highly active catalyst that can produce methanol from C 0 / H in basic medium under mild operating conditions. In the C 0 / H system, C 0 is first reduced to CO via the reverse water-gas-shift reaction. Preliminary data are presented for several metal catalysts that show activity for CO reduction. Further work is underway to select the best metal catalyst that will yield high methanol selectivity under mild conditions of temperature (< 150°C) and pressure (< 5 MPa). 2

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch011

2

2

2

2

2

2

2

2

2

Future work will extend these laboratory studies to address a key issue of H cost in the overall scheme of C 0 sequestration via C 0 hydrogénation. Here, we envision geologic formations, specifically abandoned oil and gas reservoirs, as heated slurry reactors under pressure to catalytically produce H from H 0 that is subsequently utilized to catalytically hydrogenate C 0 in these formations. Knowledge gainedfromthe laboratory studies will allow selection of active catalyst(s) that could be added to the formations to hasten reaction kinetics. This approach, if successful, holds an enormous potential for carbon sequestration. 2

2

2

2

2

2

Acknowledgments We thank Dr. Arun C. Bose of the US Department of Energy (US DOE) National Energy Technology Laboratory (NETL) for his valuable input. This work was partially funded by US DOE under Contract No. DE-AC0298CH10886.

References 1.

Carbon Sequestration: State of the Science. Office of Science and Office of Fossil Energy, U.S. Department of Energy: Washington, DC, February 1999.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

180 2. 3. 4.

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch011

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

23. 24. 25. 26.

Hindermann, J. P.; Hutchings, G. J.; Kiennemann, A. Catal. Rev.-Sci. Eng. 1993, 35(1), 1. Hennig, H. Coord. Chem. Rev. 1999, 182, 101. Ando, H.; Xu, Q.; Fujiwara, M.; Matsumura, Y.; Tanaka, M.; Souma, Y. Catal. Today. 1998, 45, 229.. Dry, M.E. Catalysis Science and Technology, vol. 1; Anderson , J. R.; Boudart, M., Eds.; Springer-Verlag, New York, 1981; p 159. Goehna, H.; Koenig, P. CHEMTECH, 1994, June, 36. Kirk-Othmer Encyclopedia of Chemical Technology (4 ed), 1991, 1, 38. Fogg, P. G. T.; Gerrard, W. Solubility of Gases in Liquids; John Wiley & Sons Ltd.: New York, New York, 1991. Yaws, C. L. Chemical Properties Handbook: McGraw Hill: New York, New York, 1999. Jessop, P.G.; Ikariya, T.; Noyori, R. Chem. Rev. 1995, 95(2), 259. Leitner, W. Angew. Chem. Int. Ed. Engl. 1995, 34, 2207. Hutschka, F.; Dedieu, Α.; Eichberger, M.; Fornika, R.; Leitner, W. J. Am. Chem. Soc. 1997, 119, 4432. Jessop, P.G.; Ikariya, T.; Noyori, R. Nature 1994, 368, 231. Tominaga, K-I.; Sasaki, Y.; Kawai, M.; Watanabe, T.; Saito, M. J. Chem. Soc. Chem. Commun. 1993, 629. Dubois, D.L. Comments Inorg. Chem. 1997, 19, 307. Lehn, J-M.; Ziessel, R., J. Organomet. Chem. 1990, 382, 157. Wegrzyn, J.E.; Mahajan, D.; Gurevich, M. Catal. Today 1999, 50, 97. Watson, R. Personal communication at Pennsylvania State University, October 26, 1999. Selley, R. C. Applied Sedimentology; Academic Press: London, 1998. Ρ 446. North, F. K. Petroleum Geology; Allen & Unwin: Boston, 1985; p 607. Chapman, R. E. Petroleum Geology; Elsevier: Amsterdam, 1983; p 415. Vairavamurthy, Μ. Α.; Schoonen, M. A. A. Geochemical transformations of sedimentary sulfur; ACS Symp. Ser.: American Chemical Society: Washington, D.C., 1995. Mango, F. D.; Hightower, J. W.; James, A. T. Nature 1994, 368, 536. Mango, F. D. Org. Geochem. 1996, 24, 977. Mahajan, D.; Vijayaraghavan, P. Fuel 1999, 78, 93. Song, C.; Murata, S., S. T.; Srinivas, T. S.; Sun, L.; Scaroni, A. W. Am. Chem. Soc.: Div. Petrol. Chem. Prepr. 1999, 44 (2), 160. th

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 12

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

Methane Dry Reforming over Carbide, Nickel-Based, and Noble Metal Catalysts Abolghasem Shamsi National Energy Technology Laboratory, U.S. Department of Energy, 3610 Collins Ferry Road, Morgantown,WV26507

Carbide catalysts of molybdenum and tungsten were prepared and tested for reaction of methane with CO at atmospheric pressure. At this pressure the catalysts were not stable and the tungsten carbide irreversibly deactivated after 35 hours on stream. The carbide catalysts produced lower H /CO ratios at lower temperatures of 650 and 750°C compared to noble metal or nickel-based catalysts. Tungsten carbide was partially oxidized to tungsten oxide during the reactions, resulting in lower catalytic activity. We also tested 1 wt% rhodium supported on alumina and two commercial Ni-based catalysts, R-67 and G-56B. Significant amounts of carbon formed on the commercial catalysts that plugged the reactor after 5 hours on stream. 2

2

182

U.S. government work. Published 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

183 Introduction Recently many researchers have concentrated their efforts toward catalytic reforming of methane with carbon dioxide. This process can be very useful for converting thermal energy into chemical energy. For example, energy losses in combustion/gasification systems and in advanced gas turbines can be captured by reacting natural gas with the by-product of combustion (C0 ) over a catalyst for producing syngas which can be converted into liquid fuels and chemicals, ultimately increasing the system efficiency. Although this concept has many environmental and economic incentives, unfortunately, there are no commercial processes for reforming of methane with C 0 (1). The main problem is that there are several carbon-forming reactions associated with this concept that deactivate the conventional steam reforming, nickel-based, catalysts. Nickel catalyzes carbon formation via hydrocarbon decomposition and CO disproportionate reactions, which greatly contributes to catalyst deactivation. Three phenomena are known to be responsible for the deactivation of nickel catalyst (2). 1) Carbon deposition, 2) metal sintering, and 3) phase transformation such as N i A l 0 * Ni/y-Al 0 . Choudhary et al. (3) reported that the pressure drop across a fixed bed reactor, containing NiO supported on CaO, increased rapidly due primarily to rapid coke formation on the catalyst surfaces. However, when they added steam to the reaction the amount of carbon was significantly reduced. Chang et.al. (4) found a coke formation rate of 7.0 wt% per hour for pentasil zeolite-supported nickel catalyst at 700°C. Furthermore, they indicated that addition of promoters such as potassium and calcium plus altering the preparation method reduced the coke formation rate to less than 0.1 wt% per hour. Addition of CaO to Ni/y-Al 0 catalyst is reported to improve the catalyst stability and also increases the reaction rate (5).

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

2

2

2

4

2

3

2

3

Operating conditions and catalyst support (6) also play an important role in catalyst particle size distribution, kinetics, and the reactivity of carbon deposited on the catalyst surfaces. Gadalla and Bower (7) recommended that a C0 /CH4 ratio of greater than one should be used for reducing carbon deposition. Furthermore, they have calculated the optimum range of temperatures for each feed composition and pressure in which carbon deposition and carbide formation are minimized. It is reported (8) that the activity and stabilty of methane reforming catalysts are significantly affected by the support and by the active metals. Reducing the concentration of Lewis acid sites on the support and reducing nickle particle size will significantly lower carbon formation. Basini and Sanfilippo (9) studied the molecular aspects of syngas production and proposed that the formation of highly reactive oxidic species, formed from breaking one 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

184 of the two C 0 bonds, is responsible for inhibiting carbon formation on the surfaces of the dry reforming catalysts such as Rh, Ru, and Ir. The effect of the support on ruthenium catalyst, under dry reforming conditions, has also been studied (10) and the catalytic activity decreases in the following order: Rh/Al 0 > Rh/Ti0 > Rh/Si0 . Several studies (11-13) have shown that the high-surface area tungsten and molybdenum carbide materials are active for methane dry reforming reactions. These catalysts appeared to be stable at higher pressures without forming significant amounts of carbon on the catalyst surfaces. However, these catalysts are very sensitive to oxidation by oxygen or H 0 , forming oxides that are not active for dry reforming. Review of the current literature indicates that developing an inexpensive catalyst, which exhibits a high selectivity toward hydrogen and carbon monoxide without forming carbon remains a challenge, specifically at higher pressure (14). In this paper, we report the preparation and testing of carbide, Ni-based and noble metal catalysts and compare catalytic activity, selectivity, and stability of these catalysts for reaction of methane with carbon dioxide. 2

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

2

3

2

2

2

Experimental Section The tungsten carbide catalyst was prepared as described (15) in U.S. patent 5321,161. The molybdenum carbide catalyst was prepared by temperatureprogrammed reduction (TPR) of molybdenum oxide in a flow of 11.6% ethane or methane in hydrogen at a flow rate of 55 ml/min. Molybdenum carbide supported on T i 0 was prepared by mixing M0O3 with T i 0 in ethyl alcohol slurry. The dried mixture was reacted with 11.6 vol.% ethane in hydrogen to form carbide. More information on carbide preparation can be found in a paper published by Lee et.al. (16). The nickel-based catalysts were prepared from water soluble nitrate solutions with proper metal ratios. A 0.3-m long quartz reactor tube (6.35-mm o.d., 4.0-mm i.d.) with a quartz thermocouple well was used as afixed-bedreactor with 0.5 grams of catalyst (-28/+48 mesh) held in place by quartz wool. Electronic mass flow controllers fed methane and carbon dioxide into the reactor, GHSV=5040 cm ^" .^ . The reactor was electrically heated to reaction temperatures. Products were analyzed by on-line gas chromatography (GC). Samples were analyzed for hydrogen, carbon monoxide, methane and carbon dioxide. A thermal conductivity detector was used with a 1-m by 3.2-mm-o.d. stainless steel molecular sieve 5A column and a 3.66-m by 3.2-mm-o.d. stainless steel HayeSep C (80/100 mesh) column at isothermal oven temperatures of 115 and 52 °C, respectively. Argon was used as carrier gas at 20 ml/min. 2

2

3

1

1

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

185 Results and Discussion A series of carbide, Ni-based, and noble metal catalysts were prepared and tested for methane dry reforming at a temperature range of 650-950°C and at atmospheric pressure. The results at 750 and 850°C are shown in Table I and the results at other temperatures are discussed in the text. Pure SiC was tested at 950°C and no significant methane or C 0 conversion was observed. We also tested pure tungsten carbide and tungsten carbide supported on silica. Methane conversion of less than 41% was obtained for the supported catalyst at 950°C. The activity of the unsupported tungsten carbide was tested at 650, 750, and 850°C. At lower temperatures of 650 and 750°C, methane and C 0 conversions were less than 15% with a H /CO ratio of 0.2. However, higher methane and C 0 conversions with a H /CO ratio of 1.1 were obtained when the temperature was raised to 850°C, at which the reaction was continued for about 45 hours as shown in Figure 1. The catalyst irreversibly deactivated after 35 hours on stream and we were not able to regenerate it with flowing hydrogen or a mixture of 11.6% ethane in hydrogen over the catalyst. Identical results were obtained when a fresh catalyst, from the same batch, was tested at a similar reaction condition (not shown). Since the freshly prepared catalyst was very sensitive to air oxidation the reactor was loaded under an argon atmosphere in a glove box. Although we were not able to determine the exact cause of the sharp decrease in catalytic activity, characterization of the sample

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

2

2

2

2

2

Table I. Methane dry reforming over carbide and Ni-based catalysts

Temperature, °C Catalysts: Mo C (me) Mo C (et) Mo C/Ti0 Tungsten carbide R-67, Topsoe* G-56B, UCI* 2

2

2

2

%CO > Conversion

750

850

750

850

750

850

750

850

18.3 36.9 69.9 5.4 94.2 96.1

92.1 89.5 70.1 90.7 98.9 99.1

37.3 59.1 87.1 12.0 91.1 90.3

99.8 99.9 88.7 99.7 95.2 97.1

37.3 51.6 76.2 13.4 95.3 93.2

96 89.4 77.8 86.6 99.1 99.8

0.3 0.7 0.9 0.2 1.0 1.0

1.0 1.1 0.9 1.1 1.0 1.0

:

(me)- TPR of Mo0 using 11.6% methane in H (et) -TPR of Mo0 using 11.6% ethane in H * Plugged the reactor after 5 hours 3

3

H /CO ratio

%CO Yield

%CH< Conversion

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

186 by X-ray diffraction after the reaction indicated that the tungsten carbide was partially oxidized to tungsten oxide, suggesting that some of the active sites had been destroyed. Molybdenum carbide catalysts were prepared by temperature-programmed reduction of molybdenum oxide with a mixture of 11.6 vol.% methane or ethane in hydrogen. The catalyst prepared with methane mixture was tested at 750 and 850°C as shown in Table I. Methane and C 0 conversions of less than 18% with a H /CO ratio of 0.3 were obtained at 650°C. At 850°C methane and C 0 conversions were higher than 90% with a H /CO ratio of 1.0. Higher catalytic activity was observed for the catalyst prepared with ethane mixture compared to that prepared with methane mixture as shown in Table I. The carbide catalysts produced low H /CO ratios and deactivated rapidly at temperatures lower than 750°C. Molybdenum carbide prepared with ethane mixture was tested for 16 hours and the results are shown in Figure 2. After 10 hours the catalytic activity continuously decreased until the reaction was terminated. This could result from oxidation of carbide active sites into the oxide, which is inactive for dry reforming reactions as reported by Claridge et. al. (13). Molybdenum carbide mixed with T i 0 was tested at 650, 750, 850, and 900°C. Methane and C 0 conversions of 21 and 42% with a H /CO ratio of 0.5 were obtained at 650°C. However, after 30 minutes on stream these numbers decreased to 9.5%, 21.8%, and 0.3, respectively. The results for the initial activity at 750 and 850°C are shown in Table I. The mixed catalyst was also unstable at temperatures of less than 850°C. Less deactivation was observed at higher temperatures and this could be due to higher stability of carbide catalyst at higher temperatures. Thermodynamically, the reduction of oxide into carbide is more favorable at higher temperatures while forming C 0 as shown in Table II. The table shows the temperatures beyond which the equilibrium constants are greater than one (K>1). At 900°C a methane conversion of 95% and C 0 conversion of 100% with a H /CO ratio of 1.0 were measured and the test was continued for 2.5 hours without a significant drop in catalytic activity or selectivity as shown in Figure 3. At temperatures of less than 850°C higher methane and C 0 conversions with a higher H /CO ratio were obtained for molybdenum carbide mixed with T i 0 than the sample without it. This shows that T i 0 has a promoting effect on the catalytic activity of the carbide catalysts. 2

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

2

2

2

2

2

2

2

2

2

2

2

2

2

2

The tungsten carbide catalyst was characterized before and after reaction using X-ray diffraction, Scanning Electron Microscopy (SEM), and X-ray Photoelectron Spectroscopy (XPS). The X-ray diffraction patterns show that the carbide catalysts were partially oxidized to W 0 during the reaction as shown in Figures 4. This is in good agreement with a study done by Clarige and his coworker (13), reporting that molybdenum and tungsten carbides were 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

187

2.0 100

1.8 1.6 1.4 1.2

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

e ο Ci

> ο a

υ

- - · - - % C02 conv.

1.0 Ο 0.8

- * - % CH4 conv.

0.6

— - A — % CO yield

0.4

w

0.2

H2/CO ratio

0.0 0

5

10 15 20 25 30 35 40 45 50 55 Time, h

Figure 1. Dry reforming of methane over tungsten carbide at 850 "C, CH4/CO2 = 1.1

2 100 1.6 80 £

c ο

Η"

60

1 Ο

VI

>

40

υ

20 Ι­

C Ο

0.8

• % C02 conv. - • • - % CH4 conv. - - A - - % CO yield - +- - H2/CO ratio

M

H 0.4

Ο

0 0

6

9

12

15

18

Time, h Figure 2. Dry reforming of methane over molybdenum carbide at 850 "C, CH/C0 = 1.16 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

188

- · — C02 conversion -at— % CO yield ••— CH4 conversion - H2/CO ratio

0.5

1

1.5

2

2.5

Time, h Figure 3. Dry reforming of methane over molybdenum carbide mixed with Ti0 , at 900 °C, CH/C0 = 1.12 2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

189

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

wc

30

40

50 60 2-Theta (deg)

70

80

Figure 4. X-ray diffraction pattern of tungsten carbide catalyst: (a) before reaction and φ) after reaction at 850 °C, CH4/C0 =1.15 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

190 Table Π. Thermodynamic calculations for Molybdenum and Tungsten Carbide Species Found in the HSC Database*

Temperature

Reaction

Equilibrium Constant (K)

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

°c 364 379 581 385 589 604 549 567 604 644

1.087 2Mo0 + 4CH4 (g) = 3C0 (g) + Mo C + 8H (g) 1.006 Mo0 + 2.5CIL, (g) = 1.5C0 (g) +MoC + 5H (g) Mo0 + 2CIL, (g) = C 0 (g) + MoC + 4 H (g) 1.050 3Mo0 + 6.5 C H 4 (g) = 4.5C0 (g) +M03C2 + 13 H (g) 1.011 1.067 2Mo0 + 3CH, (g) = 2C0 (g) + M02C + 6H (g) 1.024 3Mo0 + 5CH4 (g) = 3C0 (g) + Mo G + 10H (g) 1.032 3W0 + 2.5CH4 (g) = 1.5C0 (g) + WC + 5H (g) W 0 + 2CH4 (g) = C 0 (g) + WC + 4H (g) 1.043 1.060 2W0 + 4CH4 (g) = 3C0 (g) + W C + 8H (g) 1.16 2W0 + 3CH4 (g) = 2C0 (g) + W C + 6H (g) 3

2

3

2

2

2

2

2

2

2

3

2

2

2

2

2

3

2

2

2

3

2

2

2

2

2

2

3

2

2

2

2

2

2

2

Calculated using HSC Chemistry, Outokumpu Research Oy in Finland

converted to oxides during methane dry reforming. Further oxidation of the deactivated sample occurred when it was exposed to air. Tungsten carbides of W C and WC were detected in both samples. The SEM photographs of tungsten carbide catalyst, before and after reaction, clearly show differences between the two samples as shown in Figure 5. The particles in the sample before the reaction have sharper edges than the particles in the sample after the reaction, indicating that some of the active sites have been destroyed or covered with carbon during the reactions. The XPS analysis of tungsten carbide catalyst showed higher concentrations of carbon and oxygen on the surfaces of the sample after reaction compared to that before reaction. The ratio of carbon (C Is) to tungsten (W Is) increased from 7 to 50 while the ratio of oxygen (O Is) to tungsten (W Is) increased from 2.8 to 12.6 during 35 hours of reactions. Furthermore, no significant changes were observed for carbon (C Is) or the oxygen (O Is) peaks on these samples. The tungsten (W Is) for both samples before and after the reactions is shown in Figure 6. Two commercial nickel-based catalysts (R-67 and G-56B) were tested at 750 and 850°C as shown in Table I. Although these catalysts showed methane and C 0 conversions of more than 90% at 750°C they produced significant amounts of carbon in the catalyst bed which eventually plugged the reactor and stopped the flow as shown in Figure 7. 2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002. 2

Figure 5. SEM photographs of tungsten carbide catalyst: (a) before reaction and (b) after reaction at 850 °C, CH/C0 = 1.15

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

S ~

192

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

6500

-100

H -50

1

-45

—ι 1 1 -40 -35 -30 Electron Binding Energy (eV)

1

-25

1 -20

Figure 6. Photoelectron spectra of tungsten (W Is) carbide catalyst: (a) before reaction and (b) after reaction at 850 °C, CHyC0 =7.15 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

193

A noble metal catalyst of 1% rhodium supported on alumina was tested at 650, 750, and 850°C and the results are listed in Table HI. Methane and C 0 conversions of more than 59% with a H /CO ratio of 0.9 were obtained at 650 °C. At this temperature the catalyst showed higher activity, selectivity, and stability than carbide and Ni-based catalysts. The test was continued for more than 25 hours at 850°C, as shown in Figure 8, with no sign of deactivation or formation of significant amounts of carbon on the catalyst surface. When the space velocity (GHSV) was raised from 5040 to 7800 and 10200 cm^g^.h" no significant changes in catalytic activity or selectivity were observed. 2

2

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

1

Table ΙΠ. Methane dry reforming over 1% Rh/alumina catalyst Temperature, °C 1

GHSV.cnyV.h

H /CO ratio % CO yield % CH4 conversion % C 0 conversion 2

2

650 5040 0.9 65.6 59.2 65.2

750 5040 1.0 88 86.9 88.1

850 5040 1.0 95.7 97.2 97.4

850 7800 1.0 96.1 97.2 93.0

850 10200 1.0 94.7 95.9 92.3

The initial catalytic activity of carbide, Ni-based, and noble metal catalysts were similar at temperatures higher than 800°C. However, carbide catalysts were oxidized to oxide during the reaction and lost their catalytic activity after several days on stream. The commercial (Ni-based) catalysts formed significant amounts of carbon that plugged the reactor. Furthermore, at 650°C the initial activity of Ni-based and 1% Rh/alumina catalysts were significantly higher than the carbide catalysts. Although rhodium catalyst is much more expensive than nickel-based and carbide catalysts, it is more stable and produces little or no carbon during the reaction. Therefore, it is a good candidate for dry reforming reaction, producing syngas from methane and C 0 . The high cost of rhodium metal could be tolerated considering higher activity, low metal loading, and reduced carbon deposition. The existing nickel-based commercial catalysts are not suitable for dry reforming because significant amounts of carbon formed on the catalyst surfaces, plugging the reactor. We found that the carbide catalysts are also not reliable and deactivate irreversibly during the reaction at atmospheric pressure. However, it has been shown that these catalysts are much more stable at higher pressure (13). 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

o

ON

Si

o

I

Oo

Si

3

H

00

h-ι

H*

H*

VÔ

Os

o

o

O?

ρ

!u

κ>

α

ρ

2

τ

K>

ON

o

4t

H / C O Ratio

ON

ρ

o

s
ρ

L

ο*

o

W

• »*

4X

o

ON

Conversion/Yield (%)

00

1

il

2

ON

I

1

Or

5

3 5*

3

ON

Si

o o

ο

k>

ρ

3

ρ

o

η o o

:

«

o

+

ρ 2

ON O

H / C O Ratio

ON

bo

α

ρ

ft

S

ιι

Ο

t

o

o

00

Conversion/Yield (%)

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

o o

195 Conclusion Carbide catalysts of molybdenum and tungsten were prepared and tested for reaction of methane with C 0 . The catalysts were not stable at atmospheric pressure and irreversibly deactivated after several days on stream. The carbide catalysts produced a lower H /CO ratio than nickel-based and noble metal catalyst at lower temperatures of 650 and 750 °C. The rhodium supported catalyst was active and more stable than the carbide and Ni-based catalysts at lower temperatures of 650 and 750 °C. Significant amounts of carbon formed on the commercial catalysts, plugging the reactor after 5 hours on stream. Tungsten carbide catalyst was analyzed before and after methane dry reforming reactions using XRD, SEM, and XPS. The XRD and the XPS results showed that the tungsten carbide was partially oxidized to tungsten oxide during the reactions and may have resulted in the loss of some of the active sites. 2

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

2

Acknowledgment Author acknowledges the Natural Gas Processing and Utilization team of National Energy Technology Laboratory for funding this work. He also thanks Dr. Ranjani V. Siriwardane, James A. Poston and Elizabeth A. Frommell for carrying out the XPS, SEM and XRD measurements.

References 1.

2.

Wang, S; Lu, G. Q. Carbon Dioxide Reforming of Methane to Produce Synthesis Gas over Metal-Supported Catalysts: State of the Art. Energy & Fuels. 1996, 10, 896-904. Tomishige, K.; Chen, Y-G.; Fujimoto, K. Studies on Carbon Deposition in CO Reforming of CH over Nickel-Magnesia Solid Solution Catalysts. J. Catal. 1999, 181, 91-103. Choudhary, V. R.; Rajput, A. M. Simultaneous Carbon Dioxide and Steam Reforming of Methane to Syngas over NiO-CaO Catalyst. Ind. Eng. Chem. Res. 1996, 35, 3934-3939. Chang, J-S.; Park, S-E.; Lee, K-W.; Choi, M . J. Catalytic Reforming of Methane with Carbon Dioxide over Pentasil Zeolite-Supported Nickel Catalyst. Studies in Surface Science and Catalysis. 1994, 84, 1587-1594. Zhang, Z. L.; Verykios, X. E. Carbon Dioxide Reforming of Methane to Synthesis Gas over Supported Ni Catalysts. Catalysis Today. 1994, 21, 589-595. 2

3.

4.

5.

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

196 6.

Goula, Μ. Α.; Lemonidou, Α. Α.; Efstathiou, Α. Μ. Characterization of Carbonaceous Species Formed During Reforming of CH4 with C O over Ni/CaO-Al O Catalysts Studied by Various Transient Techniques. J. Catal. 1996, 161, 626-640. Gadalla, A. M.; Bower, B. The Role of Catalyst Support on the Activity of Nickel for Reforming Methane with CO . Chem. Eng. Sci. 1988, 43 (11), 3049-3062. Lercher, J. Α.; Bitter, J. H.; Hally, W.; Niessen, W.; Seshan, K. Design of Stable Catalysts for Methane-Carbon Dioxide Reforming. Studies in Surface Science and Catalysis. 1996, 101, 463-472. Nakamura, J.; Aikawa, K.; Sato, K.; Uchijima, T. Role of Support in Reforming of CH4 with CO over Rh Catalysts. Catalysis Letters. 1994, 25, 265-270. Basini, L.; Sanfilippo, D. Molecular Aspects in Syn-Gas Production: The CO -Reforming Reaction Case. J. Catal. 1995, 157, 162-178. York, P. E.; Claridge, J.B.; Marquez-Alvarez C.; Brungs, A. J.; Green, M.L.H. Group (V) and (VI) Transition Metal Carbides as New Catalysts for the Reforming of Methane to Synthesis Gas. Prepr. Am. Chem. Soc., Div. Fuel Chem. 1997, 42(2), 606-610. York, A. P. E.; Claridge, J. B.; Brungs, A. J.; Tsang, S.C.; Green, M.L.H. Molybdenum and Tungsten Carbides as Catalysts for the Conversion of Methane to Synthesis Gas Using Stoichiometric Feedstock. Chem. Commun. 1997, 39-40. Claridge, J. B.; York, A.P.E.; Brungs, A. J.; Marquez-Alvarez, C.; Sloan, J.; Tsang, S.C.; Green, M.L.H. New Catalysts for the Conversion of Methane to Synthesis Gas: Molybdenum and Tungsten Carbides. J. Catal. 1998, 180, 85-100. Song, C.; Srimat, S. T.; Sun, L.; Armor, J. N. Comparison of High­ -Pressure and Atmospheric-Pressure Reactions for C O Reforming of CH4 over Ni/Na-Y and Ni/Al O Catalysts Prepr. Am. Chem. Soc., Div. Petrol. Chem. 2000, 45 (1), 143-148. Vreugdenhil, W.; Sherif, F.G.; Burk, J.H.; Gadberry, J.F. U.S. Patent 5,321,161, 1994. Lee, J.S.; Oyama, S.T.; Boudart, M . Molybdenum Carbide Catalysts I. Synthesis of Unsupported Powders J. Catal. 1987, 106, 125-133. 2

2

7.

3

2

Downloaded by NORTH CAROLINA STATE UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch012

8.

9.

2

10.

2

11.

12.

13.

14.

2

2

15. 16.

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 13

A Highly Active and Carbon-Resistant Catalyst for C H Reforming with CO : Nickel Supported on an Ultra-Fine ZrO

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch013

4

2

2

Jun-Mei Wei, Bo-Qing Xu, Jin-Lu Li, Zhen-Xing Cheng, and Qi-Ming Zhu State Key Laboratory of C Chemistry & Technology, Department of Chemistry, Tsinghua University, Beijing 100084, China 1

N i supported on a specially prepared ultra-fine Z r O is studied for activity and carbon-resistant ability in catalytic reforming of CH with CO to synthesis gas. This catalyst provides a space time yield of 46.7 g CO/h·g-cat at 1030 Κ and a catalyst life longer than 600 h without detectable deactivation. Possible reasons for the extremely high activity and stability of the catalyst are discussed. 2

4

2

Introduction

The catalytic reforming of C H with C 0 for producing synthesis gas ( C H + C 0 -> 2CO + 2H ) is one of the attractive routes for utilization of the two 4

2

2

4

2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

197

198 greenhouse gases. Moreover, this process can produce synthesis gas with a H /CO ratio less than a unity, which is more desirable for Oxo syntheses, and for the syntheses of oxygenates as well as long chain hydrocarbons. Numerous papers have been documented on the catalysis of this reaction in recent years, as were reviewed by Bradford and Vannice in 1999 [1]. It is commonly recognized that Ni-based catalysts are active for this reaction, but they deactivate rapidly due to carbon deposition via CO disproportionation and/or C H decomposition [2]. Supported noble metal catalysts were found less sensitive to coking [3,4], but their practical utilization was restricted by the high cost and limited resource of the noble metals. In the search for highly active and carbon-resistant nickelbased catalysts, we showed very recently that an ultra-fine Z r 0 (u-Zr0 ) supported Ni catalyst was very active and extremely stable for the reforming reaction [5-8]. The Ni loading in the catalyst system can be as high as 27% without degrading the catalyst. Besides, we have also found that u-Zr0 itself is somewhat active for this reaction. This presentation reports on the behavior of the catalysis. 2

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch013

4

2

2

2

Experimental

The ultra-fine Zr0 with particle sizes around 6 nm and a specific surface area of 160 mVg was prepared with a precursor of drying an alcogel of ZrO(OH) under conditions for supercritical ethanol (543 K, 8.0 MPa). The alcogel was obtained by washing with anhydrous ethanol of a ZrO(OH) hydrogel prepared by addition of 0.17 M ZrOCl solution into a 2.5 M ammonia water solution with careful control of pH=9~ll. X-ray diffraction (XRD) measurement showed that the crystals of u-Zr0 were 22% monoclinic and 78% tetragonal. Ni/u-Zr0 , Ni/Al 0 , Ni/Ti0 , and Ni/Si0 catalysts were prepared by impregnating the u-Zr0 , A1 0 , Ti0 , and Si0 with an aqueous solution of nickel nitrate, followed by drying (at 383 Κ for 12 h) and calcining (at 923 Κ for 5 h). The three conventional supports were purchasedfromTianjing Institute of Chemical Industry, China. The catalytic reaction was performed under atmospheric pressure at 1030 Κ with a feed of 1:1 C H / C 0 (GHSV=2.4xl0 ml/h-g-cat). 200 mg catalysts diluted with 500 mg alfa-alumina or 200 mg supports with sizes ranging in 2040 meshes were loaded into a fixed bed quartz tubular reactor for the activity and stability measurements. Before reaction, the samples were reduced in situ at 973 Κ with flowing 10% H / N for 3 h. Ni loading in the catalysts was determined with XRF and expressed by weight percent. 2

2

2

2

2

2

2

3

2

2

2

3

2

2

2

4

4

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

199

Results and Discussions Figure 1 shows C H conversion versus reaction time on stream (TOS) over 5%Ni catalysts supported by different supports. The conversions of C H and C 0 and the selectivity to CO and H at TOS=T0 h are showed in Table 1. From figure 1 it can be seen that the initial activities over N1/AI2O3, Ni/Ti0 , and Ni/uZr0 are all quite high and only Ni/Si0 shows poor activity. Although the initial activity over Ni/u-Zr0 is not the highest, it maintains its initial activity for longer than 200 h TOS without any deactivation. In contrast, the activity of the other catalysts declines rapidly. All these observations show that supports exert significant influence on activity and stability of the Ni-based catalyst. Figure 2 shows C H conversion versus TOS over Ni/u-Zr0 catalysts with different Ni loadings. It can be seen that the catalyst activity increases with the increase of Ni loadings and that the C H conversion is close to the equilibrium value (87%) over the catalyst with a nickel loading of 27%. It can also be seen that all the Ni/u-Zr0 catalysts with different nickel loadings are very stable and that no deactivation occurs on the 27%Ni/u-Zr0 catalyst even after 600 h TOS. Besides, an obvious oscillation of C H conversion on the three catalysts takes place. Figure 3 shows C H and C 0 conversions versus TOS over u-Zr0 under conditions of atmospheric pressure, 1073 K, GHSV=2.4xl0 ml/h-g-cat. Around 10% C H conversion is obtained after an induction period and no deactivation happens after 50 h TOS. Besides, oscillation also occurs on the u-Zr0 sample. The conventional Si0 , Ti0 , and A1 0 supports with no Ni were also tested and all of them showed very low activity with C H conversion being less than 1%. The influence of space velocity (GHSV) on C H conversion over the 27%Ni/u-Zr0 catalyst was also investigated under the conditions mentioned above except with varying GHSV. Table 2 shows that increasing the GHSV results in a decrease in C H conversion, but the decreasing rate is so slow that the C H conversion can still be near the equilibrium with a GHSV as high as 4.8xl0 ml/h-g-cat indicating very high activity of the catalyst. The present 27%Ni/u-Zr0 catalyst is more active than the Nio.03Mgo.97O catalyst reported by Fujimoto et al. [9]. C H conversion over the latter catalyst is 82% at 1123 Κ and GHSV=1.8xl0 ml/h-g-cat (W/F=1.2 hg-cat/mol), corresponding to a STY of 11.9 g CO/hg-cat. For comparison, the present catalyst provides 85% C H conversion at 1030 Κ (ca.100 Κ lower than over Nio.03Mgo.97O) and GHSV=4.8xl0 ml/h-g-cat, corresponding to a STY of 46.7 g CO/h-g-cat. The extremely high activity and stability of the Ni/u-Zr0 catalysts may be accounted for with the unique character of the u-Zr0 and the structure of the catalyst. On one hand, the u-Zr0 support is able to accommodate as high as near 30% Ni without degrading the catalyst. This is the reason why our catalyst is much more active than Fujimoto's, in which only 4.26 wt%Ni was contained. 4

4

2

2

2

2

2

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch013

2

4

2

4

2

2

4

4

2

2

4

4

2

2

2

2

3

4

4

2

4

4

4

2

4

4

4

4

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

200

80

ο

fi ο

0

-—-ο— —

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch013

Μ

u > fi ο

—^—Ni/u-Zr02 —*--Ni/AI203 —•—-Ni/Si02

20

υ

-A—Nî/Ti02 •1

•

1

1

50

100

150

200

Time on Stream / h Figure 1: CH conversion versus reaction time over Ni supported on different support 4

9 0 s©s

0 73

80

Ο

u

70

- 5%Ni/u-Zr02

> Ο

ο

ta u

- 125%Ni/u-Zr02 60t -27%Ni/u-Zr02 5 0

H10 0

J

1 20 0

1

1 300

1

1 400

500

6 00

Time on stieam / h Figure 2: CH conversion versus reaction time over Ni/u-Zr0 catalysts with different Ni loadings 4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002. 4

2

4

79.5 87.9 83.2 91.1 86.6

95.4 96.7 97.3 98.4 95.2

1.20 1.10 1.23 1.08 1.10

1.0 1.5 2.0 2.8 3.7

88.3 86.7 86.4 81.2 77.5

86.2

85.9

85.8

79.7

79.1

2.4

3.6

4.8

7.2

9.6

Reaction temperature: 1030 K; The data represent the results at TOS=50 h.

83.4 97.6 1.17

H Sel. /% CO Sel. /% CO/H2

0.5

1

100.0

4

CH TOF / s"

2

C 0 Conv. /%

88.7

4

C H Conv. /%

1.2

GHSV / 10 ml-h-g-cat

2

Table 2 Effect of GHSV on the activity and selectivity over 27%Ni/u-Zr0 catalyst

4

2

85.9

98.7

1.18

1.1

67.8

Reaction conditions: T=1030 K , P=0.1 M P a , G H S V = 2 . 4 x l 0 ml/g-cat-h,C0 /CH =l

': The data represent the results at TOS=10 h

Ni/u-Zr0

2

65.0

95.7

97.8

1.02

0.5

16.4

11.6

2

Ni/Si0

84.2

87.7

1.05

2

2.5

66.5

67.0

2

Ni/Ti0

2

85.5

CO/H 90.6

4

1.06

2.9

71.3

3

70.9

2

,a

Νί/γ-Α1 0

2

TOF C H /s-

a

C 0 Conv. /%

4

CH Conv. /%

Catalyst

H Sel. /%

2

C O Sel. /%

4

Table 1 Activity and selectivity of Ni-based catalysts for C H reforming with C 0 over different Ni-based catalysts

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch013

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch013

202

Figure 3: CH and C0 conversions versus reaction time over u-Zr0 catalysts 4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

203

In our Ni/u-Zr0 catalyst, Ni particles are dispersed among u-Zr0 , there is no limitation to the loading of Ni, though there may be some interaction between Ni and u-Zr0 . In fact, TEM and other characterization data show that the Ni/uZ r 0 catalyst can better be described as a nano-composite composed of nano-Ni metal and nano-Zr0 crystals. It is assumed that the nano-composite nature of the catalyst is essential for the long-lasting anti-carbon property of the catalyst. On the other hand, the fact that u-Zr0 itself possesses some activity for the desired reaction implies that, though being a support, it can activate both C H and C 0 . This would appreciably increase the number of the active sites and therefore enhance the activity. In addition, surface oxygen formed by the dissociation of C 0 on u-Zr0 at the interface between metal and support [5] facilitates the elimination of carbon, thereby the stability of the catalyst is enhanced. This mechanism could be effective for the present Ni/u-Zr0 catalyst because Ni and u-Zr0 particles are very tiny and therefore they are intimately dispersed and contacted. The large area of the interface and the close distance between Ni and u-Zr0 allow the surface oxygen to migrate onto the Ni surface and react with carbon easily. Oscillation may stem from the alternative generation and elimination of carbon. Further investigations on this phenomenon are being made. 2

2

2

2

2

2

4

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch013

2

2

2

2

2

2

Acknowledgement

This work was financially supported by the grants from the Fundamental Research Foundation of Tsinghua University and National Natural Science Foundation of China (NSFC, grant 0094 ).

References

1. 2. 3. 4.

Bradford, M.C.J., Vannice, M.A., Catal. Rev.-Sci. Eng., 1999, 41, 1. Claridge, J.B., Green, M.L.H., Tsang, S.C., York A.P.E., Ashcroft A.T., Catal. Lett., 1 9 9 3 , 22, 299. Ashcroft, T., Cheetham, A.K., Greenand, M.L.H., Vernon, P.D.F., Nature, 1 9 9 1 , 352, 225. Bitter, J.H., Seshan, K., Lercher, J.A., J. Catal., 1 9 9 8 , 176, 93.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

204 5. 6. 7. 8.

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch013

9.

Wei, J.M., Xu, B.Q., Li, J.L., Cheng, Z.X., Zhu, Q.M., Appl. Catal. A: General, 2000, 196,L167. Wei, J.M., Xu, B.Q., Cheng, Z.X., Li, J.L., Zhu, Q.M., Stud. Surf. Sci. Catal., 2000, 130D, 3687 Wei, J.M., Xu, B.Q., Li, J.L., Cheng, Z.X., Zhu, Q.M., Fuel Chemistry Division preprint, 2001, 46, xx Xu, B.Q., Wei, J.M., Wang H.Y., Sun K.Q., Zhu, Q.M., Catal. Today, in press Tomishige, K., Yamazaki, O., Chen, Y., Yokoyama, K., Li, X., Fujimoto K., Catal. Today, 1998, 45, 35.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 14

CO Reforming of Methane over Ru-Loaded Lanthanoid Oxide Catalysts 2

Kiyoharu Nakagawa , Shigeo Hideshima , Noriyasu Akamatsu , Na-oko Matsui , Na-oki Ikenaga , and Toshimitsu Suzuki 1,2

1

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

1

1

1

1,2,*

1

2

Department of Chemical Engineering, Faculty of Engineering, and High Technology Research Center, Kansai University, Suita, Osaka 564-8680, Japan

CO reforming of methane was carried out using ruthenium-loaded catalysts. The productivity of the synthesis gas from methane was strongly affected by support materials and pretreatment conditions. At 600 °C, methane conversion and yields of CO and H were the highest with the Ru/Y O . CO pretreatment at the reaction temperature were effective for the activation of Ru-loaded La O and Y O catalysts. During CO pretreatment, loaded ruthenium species were transformed into metallic Ru and RuO . Coexistence of metallic Ru and RuO might be important to exhibit high catalytic activity for the CO reforming of methane. 2

2

2

2

3

2

3

2

3

2

2

2

2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

205

206 The production of synthesis gas from methane has been studied in three reactions that attract industrial interest; the partial oxidation of methane (reaction 1), steam reforming (reaction 2), and C 0 reforming (reaction 3). 2

CH + l/20 4

2

C O + 2 H Δ H° = -36 kJ/mol (1) 2

298

C H +H 0 ^ C O + 3 H Δ H °

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

4

2

2

= 2 9 8

C H + C 0 —> 2CO + 2 H Δ H 4

2

0

2

+206 kJ/mol (2)

^ +247 kJ/mol (3)

Among those, the C0 -reforming of methane has recently attracted much attention (reaction 3) (1-3). C 0 reforming of methane to synthesis gas was first reported by Fischer and Tropsch (4). This reaction can also contribute to the utilization of carbon dioxide as carbon sources. Since C 0 and C H are greenhouse gases, their conversion to synthesis gas could be a method of its reduction from the environment (5). The transition metals of the group VIII loaded on metal oxides have been widely examined as catalysts for the C 0 reforming of methane (6, 7), and in some cases the effect of the support on the catalytic activity has been reported. Rostrup-Nielsen et al. have shown the order of reactivity for C 0 reforming to be Ru > Rh > N i , Ir > Pt > Pd. This order is similar to the proposed order for steam reforming, but the superiority of Rh and Ru is less pronounced (8). Solymosi et al have reported the order of reactivity is Ru > Pd > Rh > Pt> Ir, when A 1 0 was used as a support (9). Zhang et al reported the effect of support on the Rh catalysts and concluded that the activity order is Y S Z (yittria stabilized zirconia) > A1 0 > T i 0 > S i 0 > L a 0 > MgO (10). Nakamura et al. have investigated a significant effect of the support on the catalytic activity over Rh-loaded catalysts and the activity order is reported to be A 1 0 > T i 0 > S i 0 (11). With the Pd catalysts, the activity order is reported to be T i 0 > A 1 0 > Si0 >MgO(72;. Carbon formation over metal catalysts during the C 0 reforming has been observed (13, 14). The deposited carbon is formed via different routes (15). The conventional N i based catalysts for steam reforming tend to be coked due to the formation of stable nickel carbide on the surface of the catalyst (16). Zhaolong et al. (16, 17) reported that N i / L a 0 catalyst exhibited high activity and longterm stability. On N i / L a 0 catalyst, the rate increased during the initial 2-5 h of reaction and then became stationary value with time on stream. An X-Ray diffraction study revealed that a large amount of C 0 was stored in the support as L a 0 C 0 . Fujimoto et al. (3, 18) have found that Nio.03Mgo.97O and Nio.03Cao.13Mgo.84O solid solution catalysts which were reduced with hydrogen a 2

2

2

4

2

2

2

3

2

3

2

2

2

3

2

3

2

2

2

2

2

2

3

3

2

2

2

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

2

3

207 high temperature, showed exce lient activities and stabilities in a run for 100 hat 850 °C under 3.2 atm of total pressure and CH /C0 = 1/1. We have reported in previous papers that the performance of iridium-loaded catalysts depended strongly upon the support materials in the methane to synthesis gas (19) and that the interaction between C 0 and the support strongly affects the reactivity of Ir-loaded catalysts in the C 0 reforming of methane (20). In the C0 reforming of methane and heptane, Ru/La 0 catalyst exhibited high activity (21-23). To maintain a higher catalytic activity over Ru/La 0 catalyst, the formation of La 0 C0 seemed to be essential. The interaction between C 0 and the support over Ru-loaded catalysts in the pulsed reaction was discussed in previous papers (22, 23). In this paper, in order to exploit more active and stable catalyst for C 0 reforming of methane we will discuss effect of support on the C 0 reforming of methane over Ru loaded catalysts. It also describes studies ofmarked promoting effect of C 0 pretreatment on the C 0 reforming of methane over rare earth oxides-supported Ru catalysts. 4

2

2

2

2

2

3

2

2

2

3

3

2

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

2

2

2

2

Experimental The catalyst supports used were A1 0 (JRC-ALO-4 the reference catalyst provided by the Catalyst Society of Japan), Si0 , CaO (Wako Pure Chemical Industries. Ltd.), MgO (Ube Industry Co. Ltd.), Ti0 , ΖιΌ (Japan Aerosil Co.), Y 0 , La 0 , and Nd 0 which were obtained by thermal decomposition of Y (C 0 ) -4H 0 (Wako Pure Chemical Industries. Ltd.), La(CH COO) -1/2H 0, and Nd (C0 ) (Nacalai tesque, Inc.), Ce0 (Nacalai tesque, Inc.), Sm 0 , Eu 0 (Anan Kasei. Co.), Pr O , Gd 0 , Tb 0 (Santoku Kinzoku Industries Ltd.). The supported catalysts containing 0.5 wt%Ru metal were prepared by impregnation methods with an aqueous solution o f R u C l H 0 (Mitsuwa pure chemicals) onto suspended supports, followed by evaporation-to-dryness. Loaded catalysts were calcined at 600 °C for 5 h in air. Prior to the reaction, the catalyst was pretreated with CH , H , or C 0 at 400 or 600 °C for 1 h. The reaction was carried out with a fixed-bed flow-type quartz reactor(300 χ 8 mm) at 1 atmosphere pressure. The conditions for C 0 reforming, using 100 mg of a catalyst and 300 mg of silica sand (Merck), 30 mL/min of C H and 30 mL/min of C0 were introduced at 600 °C. Silica sand was used as a heat buffer for a large endothermic reforming reaction. The reaction products (H , CO, C H , and C0 ) were analyzed by a gas chromatograph directly connected to the outlet of the reaction tube with a thermal conductivity detector (M200 chromato analyzer, Nippon Tyran Co.), which can separate H , CO, C H , and C 0 , within a few hundred seconds. The specific surface areas of catalysts were measured with the BET method using N at -196 °C with an automatic Micromeritics Gemini model 2375. 2

3

2

2

2

2

3

2

2

4

3

2

3

2

2

3

3

3

2

6

n

2

3

4

2

3

2

2

3

7

3

4

2

3

2

2

2

4

2

2

2

4

4

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

3

208 Powder X-ray diffraction (XRD) patterns were measured by a Simadzu X R D 6000 using monochromatized Cu-K α radiation. Results and discussion Table I and Figure 1 show the product distributions and effect of time on stream on the C H conversion over various metal oxide Ru-loaded catalysts. In all cases, the C 0 conversion was higher than that of C H . The support activity order at 600 °C was Y 0 è L a 0 è N d 0 >Pr 0,, ^ S m 0 ê S c 0 ^ G d 0 > C e 0 > M g O ^ T i 0 > CaO è T b 0 , A1 0 > S i 0 . The performance of the catalyst depended strongly upon support materials. A t 600 °C, methane conversion and yields of CO and H were the highest with the Y 0 -supported Ru catalyst, providing an H to CO ratio of 0.83. L a 0 and N d 0 supports also exhibited high activities at 600 °C. Characteristic features of the rare earth oxide supports exhibited relatively high activities irrespective of the surface area. In addition, no carbon deposition was observed. C H conversions gradually increased with time-on-stream over Ru-loaded Y 0 and L a 0 catalysts. On other supports, C H conversions gradually decreased with time on stream. The significant effect of the support might be ascribed to the activation o f C 0 with metal oxides used as a support. Similar results have been obtained in the C 0 reforming of heptane using a Ru-loaded catalyst (21). To achieve a higher activity in t h e C 0 reforming, the activation of C 0 on the support would be one important factor, in addition to the activation of methane on the transition metal surface. Nakamura et al (11) have reported the effects of supports on catalytic activities in the C 0 reforming of methane. They obtained an enhanced catalytic activity by mixing M g O with Rh/Si0 , concluding that promoted activity is ascribed to the dissociation of C 0 on the surface of Rh enriched with the C 0 adsorbed on MgO. This suggests that C 0 dissociation is the rate-determing step of this reaction. C H conversion with A1 0 , T i 0 , and Si0 -supported catalyst decreased gradually with increasing reaction time, presumably due to the accumulation of carbonaceous material on the catalyst surface. Carbon deposition through the Boudouard reaction (reaction 4) or methane decomposition (reaction 5) is thermodynamically favorable below 900 °C (15, 24). 4

2

4

2

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

2

3

3

2

2

3

2

2

3

4

6

7

2

2

3

3

2

2

2

2

2

3

2

3

2

3

3

4

2

3

2

3

4

2

2

2

2

2

2

2

2

2

4

2

3

2 C O ^ C + C0

CH ^ 4

2

ΔΗ°298=-

2

C + 2H

2

2

1 7 3 k J / m o l (4)

A H ° 2 9 8 = + 7 5 k J / m o l (5)

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

209 Table L Effect of supports of Ru(0.5 wt%) catalyst on the CO2 reforming of methane at 600 °C Catalyst S.A. Conversion Yield H2/CO H2 CO CH4 C02 (ratio) (m /g) (%) (%) 0.83 Y203 32.7 12.9 29.9 35.5 27.1 0.83 30.5 La203 23.7 28.0 25.4 33.0 0.72 31.6 Nd203 4.6 22.7 27.2 36.1 0.73 26.3 19.3 ΡΓ6Ο11 9.9 22.8 29.7 0.70 S1112O3 17.7 25.4 3.2 21.6 29.2 0.68 23.0 SC203 27.1 15.7 19.3 26.6 0.62 Gd203 22.1 17.8 26.3 13.6 1.4 0.56 19.2 Ce02 50.7 10.8 15.0 23.4 0.54 16.4 8.9 MgO 32.1 12.7 20.2 0.66 15.5 T1O2 41.2 18.7 9.2 12.3 0.50 12.6 6.3 CaO 4.6 9.5 15.8 0.38 13.4 Tb407 13.5 5.1 9.3 17.5 0.24 11.6 Yb203 2.8 2.8 7.2 16.0 0.28 A1203 3.0 10.5 174.0 6.8 14.3 0.18 8.0 Si02 151.6 4.7 1.4 11.3 Catalyst, 0.100g; Silica sand, 0.300g; Space velocit^o^OOh^mL g-caf Flow rate, 60.0mL/min(CH4/CO2=1.0); Reaction time = 2 h

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

2

1

-

0

-

Y 0 2

3

-a— - 0 — Ce0

I

2

- - 0 - - MgO Ti0 2

A1.03 I

k

II 40

60

Si0

2

80

Time-on-stream (min) Figure 1. Effect of time-on-stream on the conversion of CH4 for CO 2 reforming of methane over Ru loaded catalysts. Reaction temperature = 600 °C , Catalyst = 100 mg + 300 mg of silica sand, Ru loading level = 0.5 wt%, Flow rate = 60.0 mL/min (CHVC0 - 1.0), Space velocity = 36000 Η ml/g-catalyst 2

1

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

210

Kinetically, both the Boudouard reaction (reaction 4) and the methane decomposition reaction (reaction 5), which give undesirable carbon, are known to be exceptionally slow in the absence of a catalyst, but both can be readily catalyzed by many transition metals. Figure 2 shows X R D analyses of Ru(5.0wt%)/La O catalyst after the reaction. Lanthanum oxide was transformed into La 0 C03 during reaction. Similar results were reported (16, 21). In addition, metallic Ru and R u 0 were detected. This result would suggest that to exhibit high activity for C 0 reforming of methane over Ru/La 0 , formation of L a 0 C 0 and coexistence of metallic Ru and R u 0 were important. In Ru-loaded Y 0 and L a 0 catalysts, C H conversions gradually increased with time on stream. To investigate increased C H conversions, effect of pretreatment on the C 0 reforming of methane was carried out with Ru loaded catalysts, and the relations between the activity and product distributions were examined. Figures 3 and 4 show the effect of various pretreatment conditions on the catalytic activity over Ru/La 0 and R u / Y 0 catalysts. H reduction, C 0 pretreatment, and C H pretreatment were carried out at 600 °C for lh, respectively. H reduction followed by C 0 pretreatment was carried out, at first H reduction for lh, following C 0 pretreatment for lh at 600 °C. Similar results were obtained using L a 0 and Y 0 supports. When Ru/La 0 and R u / Y 0 catalysts were pretreated with C H , they were not these catalysts activated for C 0 reforming of methane. C 0 pretreatment or H reduction were effective pretreatments for L a 0 and Y 0 supports. Most effective pretreatment was H reduction followed by C 0 pretreatment. Table II shows effect of pretreatment on the activity for the C 0 reforming of methane over various metal oxide supported Ru catalysts. C H and C 0 conversions were greatly affected by the support materials. In all the cases, the C 0 conversion was higher than that of C H . In L a 0 , Y 0 , T i 0 , and S i 0 supported cases, H yield and H /CO ratio after C 0 pretreatment were higher than that without pretreatment. At 600 °C, all the Ru-loaded catalysts without C 0 pretreatment afforded a lower H /CO ratio than the stoichiometric ratio of unity. Low H / C O ratios over Ru-loaded catalysts could be attributed to the reverse water-gas shift reaction shown below: 2

2

3

2

2

2

2

3

2

2

3

2

2

3

2

3

4

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

4

2

2

2

3

2

3

2

2

2

2

2

2

2

4

3

2

3

3

2

3

4

2

2

2

2

3

2

2

3

2

2

4

2

4

2

2

2

3

2

3

2

2

2

2

2

2

2

C 0 + H ^ CO + H 0 Δ Η 2

2

2

0 298

= +41kJ/mol (6)

The excess C O may come from this reaction, and H was consumed to increase C 0 conversion. This reaction seems to be faster than the C H - C 0 reaction. In contrast, in cases of C 0 pretreatment, H /CO ratio was remarkably improved. It seemed that CO? pretreatment suppressed the reverse water-gas shift reaction. 2

2

4

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

211

Figure 2.XRD pattern of Ru(5.0 wt%)/La 0 wt%)/La 0 was reacted at 600 °C for lh. :

2

3

catalyst after reaction. Ru(5

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Τ3 ?

-α

g

ο

Ο

g

h*

ΧΛ

ϊ

•—ν

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

U

0.2

0.4

20

60

-L 80

Time-on-stream (min)

J40 100

2

120

1

2

4

- C H pretreatment

2

• - O - - H , C 0 pretreatment

2

—• — C 0 pretreatment

- £ r — ^reduction

— Ο — Without pretreatment

2

3

2

Figure 3. Effect of pretreatment gases on the activity of C0 reforming of methane over Ru/La 0 catalyst Reaction temperature = 600 °C, Catalyst = WO mg + 300 mg of silica sand, Ru loading level = 0.5 wt%, Flow rate = 60.0 mL/min (CHJCQ = 10), Space velocity = 36000 H ml/g-catalyst

ο

©

0.8

1

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

υ

Ο

g

ï

— / s

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

0

0.2

0.4

0.6

20

JX 40 60 80 Time-on-stream (min) 100

120

2

2

l

4

- C H pretreatment

2

• - Ο - - Η , C 0 pretreatment 2

— • — C 0 pretreatment

- Δ — Η reduction

— Ο — Without pretreatment

2

2

2

3

Figure 4. Effect ofpretreatment gases on the activity of C0 reforming of methane over Ru/Y 0 catalyst. Reaction temperature = 600 °C, Catalyst =100 mg + 300 mg of silica sand, Ru loading level = 0.5 wt%, Flow rate = 60.0 mL/min (CH/C0 = 1.0), Space velocity = 36000 h~ ml/g-catalyst

Ο ρ

i

0·

0.8

1

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

in

fcs>

216

Table Π. Effect of pretreatment on the activity for CO2 reforming of CH4 at 600 C over Ru(0.5wt%)-loaded various supports Catalyst Reaction Conversion Yield H2/CO time CH4 C02 H2 CO (min) (ratio) (%) (%) La2U3 5 22.2 29.1 18.8 25.7 0.73 Without pretreatment 60 27.3 33.1 24.4 30.2 0.81 120 28.0 33.0 25.4 30.5 0.83 La2Ô3 5 25.5 29.5 23.6 27.6 0.86 C02-pretreatment 60 30.4 33.7 28.8 32.1 0.90 120 30.9 33.7 29.5 32.3 0.91 Y2O3 5 20.2 27.1 16.8 23.7 0.71 Without pretreatment 60 28.8 34.0 26.2 31.4 0.83 120 29.9 35.5 27.1 32.7 0.83 Ϋ2Ο3 5 21.7 25.6 19.7 23.6 0.83 C02-pretreatment 60 28.4 31.6 26.8 30.0 0.89 120 28.9 31.2 27.8 30.0 0.93 AI2O3 5 6.8 13.2 10.0 3.6 0.37 Without pretreatment 60 8.4 15.9 4.7 12.2 0.39 120 6.8 14.3 3.0 10.5 0.28 ΑΓ2Ο3 5 7.1 11.8 4.7 9.5 0.49 C02-pretreatment 60 5.5 11.6 2.5 8.6 0.29 120 4.2 10.1 1.3 7.2 0.18 T1O2 5 16.5 23.1 13.3 19.8 0.77 Without pretreatment 60 14.3 21.1 11.0 17.7 0.75 120 12.3 18.7 9.2 15.5 0.77 T1O2 5 26.5 29.1 25.2 27.8 0.91 C02-pretreatment 60 20.9 23.9 19.5 0.87 22.4 120 17.6 21.4 15.8 19.5 0.81 S1O2 5 8.8 16.1 5.1 12.5 0.38 Without pretreatment 60 6.0 13.0 2.5 9.5 0.47 120 4.7 11.3 1.4 8.0 0.43 S1O2 5 17.2 19.9 15.9 18.6 0.85 C02-pretreatment 60 8.7 13.7 6.3 11.2 0.56 120 6.3 3.9 11.1 8.7 0.45 Flow rate, 60.0mL/min(CH4/CO2=1.0); Reaction time = 2h Catalyst, 0.10g; Silica sand, 0.300g; Space velocity=36,000h mL g-cat

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

0

1

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

1

217 Figures 5, 6, and 7 show X R D analyses of Ru(5.0wt%)/Y O and Ru(5.0wt%)/Al O catalysts before and after A r or C 0 pretreatments. A l l catalysts were calcined in air at 600 °C for 5h before the pretreatment Therefore, at first, R u 0 was formed in the major phase of Ru species. In the R u / Y 0 catalyst, the chemical species of ruthenium after C 0 pretreatment were observed both metallic ruthenium and R u 0 . Co-existence of Ru metal and Ru oxide were more effective for C 0 reforming of methane than single species of metallic Ru or Ru oxide (see Fig. 2). It seems that co-existence of metallic Ru and Ru oxide would play important roles for C 0 reforming of methane. Thus higher catalytic activity would be exhibited after C 0 pretreatment over R u / Y 0 catalyst. After Ar pretreatment, coexistence of Ru metal and Ru oxide was not detected. In the R u / Y 0 catalyst, C 0 pretreatment greatly enhanced formation of metallic Ru. During the CO? pretreatment, C O was observed, and C O might react with R u 0 to give metallic Ru. In contrast, in the R u / A l 0 catalyst, where catalytic activity was not improved by C 0 pretreatment, Ru metal was not detected after C 0 pretreatment. Therefore, higher catalytic activity would not be exhibited. In addition, X-ray photoelectron spectroscopic (XPS) analysis also indicated that ruthenium species of the pretreated R u / Y 0 catalyst was found to be more reduced state than the untreated R u / Y 0 catalyst. Previously, C 0 was behaved to be an oxidant of several lower valent transition metal oxides (Fe 0 /activated carbon (25), V 0 / M g O (26)). Such, reducing effect of C 0 is not consistent with previous results. However, such reducing effect of C 0 was only observed over L a 0 and Y 0 supports, which were basic oxides and were partially transformed into carbonate, to give strongly adsorbed CO? species. Electron transfer from these species might have affected oxidation state of ruthenium species. Table III lists effect of pretreatment on the activity for the C 0 reforming of methane over noble metal-loaded Y ? 0 catalysts. Except Ir case, the C H conversion, synthesis gas yield and H / C O ratio after C 0 pretreatment were higher than that of without pretreatment. Especially, in the case of R h / Y 0 catalysts, the C H , C 0 conversion, and synthesis gas yield remarkably increased. To improve a higher activity in the C 0 reforming of methane, the activation of C 0 on the Y ? 0 support would be one of important factors, in addition to the modification of surface metal spices. 2

2

3

3

2

2

2

3

2

3

2

3

2

2

2

2

2

2

3

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

2

3

2

2

2

2

2

3

2

3

4

2

4

2

2

2

3

2

3

2

3

4

2

2

2

4

3

2

2

2

3

Conclusions C 0 reforming of methane with supported Ru catalysts was strongly affected by support oxides and pretreatment. At 600 °C, Ru/Y?0 catalyst showed the highest catalytic activity without pretreatment. C 0 pretreatment or H reduction were effective pretreatment for La?0 and Y 0 supported Ru catalysts. During C 0 pretreatment, co-existence of metallic Ru and R u 0 were 2

3

2

2

3

2

3

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

218

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

a) Fresh

Ο Ru02 V Ru

b) CO2 pretreatment 600 °C lh

30

35

40

45

50

55

60

20/deg. (CuKa) Figure 5. XRD patterns of Ru(5.0 wt%)/Y:0 catalyst before and after heat treatment under CO 2. a)Ru(5.0 wt%)/Y'2O3 was calcined at 600 °C in air. b) CO2 pretreatment at 600 °C for lh. 3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

219

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

a) Fresh

Ϊ 30

Ο

1 35

1 40

1

1

45

50

Ru02

1 55

1 60

2 0/deg. (CuKa) Figure 6. XRD patterns of Ru(5.0wt%)/Y O catalyst before and after heat treatment under Ar. a)Ru(5.0 wt%)/Y 0 was calcined at 600 °C in air. b) Ar pretreatment at 600 °C for lh. 2

2

3

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

220

a) Fresh

Ο Ο

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

ORu02 • AI2O3

CO

S

b) CO2 pretreatment 600 °C lh

Ο

Ο

•

20

40

2 0/deg. (CuKa) Figure 7. XRD patterns ofRu(5.0wt%)/Al O catalyst before and after heat treatment under CO 2. a)Ru(5.0 wl%)/Al 0 was calcined at 600 °C in air. b) CO: pretreatment at 600 °C for lh. 2

2

3

}

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

221

Table ΠΙ. Effect of pretreatment on the activity for CO2 reforming of CH4 at 600 C over variours metals-loaded Y2O3 catalysts Catalyst Reaction Conversion Yield H2/CO time CH4 CO2 H2 CO (min) (ratio) (%) (%) Ru(0.5wt%)/Y2O3 5 20.2 27.1 16.8 23.7 0.71 Without pretreatment 60 28.8 34.0 26.2 31.4 0.83 120 29.9 35.5 27.1 32.7 0.83 Ru(ÔJwt%)/Y203 5 21.7 25.6 19.7 23.6 0.83 60 C02-pretreatment 28.4 31.6 26.8 30.0 0.89 120 28.9 31.2 27.8 30.0 0.93 Rh(0.5wt%)/Y2O3 5 18.9 23.8 16.4 21.3 0.77 Without pretreatment 60 19.5 25.0 16.8 22.3 0.75 120 18.1 22.6 15.8 20.4 0.77 5 31.2 Rh(0Jvit%j7Y2O3 35.1 29.3 33.1 0.89 C02-pretreatment 60 31.2 35.5 29.1 33.4 0.87 120 30.1 33.7 28.3 31.9 0.89 Ft(0.5wt%)/Y2O3 5 18.9 23.8 16.4 21.3 0.77 Without pretreatment 60 19.5 25.0 16.8 2.3 0.75 120 18.1 22.6 15.8 20.4 0.77 Κ(0.5^ί%)/Ϋ2θ3 5 21.6 26.2 19.3 23.9 0.81 60 C02-pretreatment 20.6 24.7 18.5 22.7 0.81 120 19.9 24.4 17.7 22.2 0.80 lr(0.5wt%)/Y2O3 5 11.0 18.4 7.3 14.7 0.50 Without pretreatment 60 9.1 16.4 5.5 12.8 0.43 120 8.5 15.9 4.8 12.2 0.39 5 Ir(0:5wt%)7Y2O3 8.7 16.0 4.9 12.3 0.40 60 C02-pretreatment 8.0 15.9 4.1 11.9 0.34 120 7.7 15.4 3.8 11.5 0.33 Pd(0.5wt%)/Y2O3 5 8.4 16.0 4.6 12.2 0.38 Without pretreatment 60 8.9 15.3 5.7 12.1 0.47 120 7.5 13.5 4.5 10.5 0.43 5 Μ(0:5Μ%)7Ϋ2θ3 14.0 21.5 10.3 17.8 0.58 60 C02-pretreatment 12.1 19.1 8.6 15.6 0.55 120 10.6 17.7 7.1 14.1 0.50 Flow rate, 60OmL/min(CH4/CO2=1.0); Reaction time = 2h Catalyst, 0.10g; Silica sand, 0.300g; Space velocity^MOOh^mL g-cat 1

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

222 detected. In order to achieve a high catalytic activity, co-existence of metallic Ru and R u 0 might be important. 2

Acknowledgements

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

This work was has been carried out as a research project of The Japan Petroleum Institute commissioned by the Petroleum Energy Center with the subsidy of the Ministry of International Trade and Industry, Japan. K. Nakagawa is grateful for his fellowship from the Japan Society for the Promotion of Science (JSPS) for Young Scientists. References 1. Richardson, J. T.; Paripatyadar, S. A . Appl. Catal. 1990, 61. 293. 2. Ashcroft, A. T.; Cheetham, A . K.; Green, M . L. H.; Vernon, D . D. F. Nature. 1991, 352. 18. 3. Tomishige, K.; Fujimoto, K . Catal. Surv. Jpn. 1998, 2. 3. 4. Fischer, F.; Tropsch, H. Brennstoff Chem. 1928, 9. 39. 5. Rostrup-Nielsen, J.R. Stud. Surf. Sci. Catal. 1994, 81. 25. 6. Sodwsawa, T.; Dobashi, Α.; Nozaki, F.; React. Kinet. Catal. Lett. 1979, 12. 107. 7. Erdöhelyi, Α.; Cserényi, J.; Solymosi, F. J. Catal. 1993, 141. 287. 8. Rostrup-Nielsen, J.R.; Hansen, J.H.B. J. Catal. 1993, 144. 38. 9. Solymosi, F.; Kutsan, G.; Erdöhelyi, A . Catal. Lett. 1991, 11. 149. 10. Zhang, Z.L.; Tsipouriari, V . A . ; Efstathiou, A . M . ; Verykios, X . E . J. Catal. 1996, 158. 51. 11. Nakamura, J.; Aikawa, K.; Sato, K.; Uchijima, T. Catal. Lett. 1994, 25. 265. 12. Erdöhelyi, Α.; Cserényi, J.; Papp, Ε.; Solymosi, F. Appl. Catal. A. 1994, 108. 205. 13. Vernon, P.D.F.; Green, M . L . H . ; Cheetham, A . K . ; Ashcroft, A . T . Catal. Today. 1992, 13. 417. 14. Rostrup-Nielsen, J.R. Catal. Today. 1993, 18. 305. 15. Rostrup-Nilsen, J.R. in Catalysis, Science and Tecnology, edited by Anderson, J.R.; Boudart, M . Springer, Berlin, 1984; Vol. 5, pp 1-117. 16. Zhang, Z . L . ; Verykios, X . E . Appl. Catal. A. 1995, 138. 109. 17. Zhang, Z.L.;Verykios, X.E.;MacDonald, S . M . ; Affossman, S. J. Phy. Chem. 1996, 100. 744. 18. Yamazaki, O.; Nozaki, T.; Omata, K.; Fujimoto, K. Chem. Lett. 1992, 1953. 19. Nakagawa, K . ; Ikenaga, N . ; Suzuki, T.; Kobayashi, T.; Haruta, M. Appl. Catal. A. 1998, 169. 281. 20. Nakagawa, K . ; Anzai, K . ; Matsui, N.; Ikenaga, N.; Suzuki, T.; Teng, Y . ; Kobayashi, T.; Haruta, M . Catal. Lett. 1998, 51. 163.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

223 21. 22. 23. 24. 25.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch014

26.

Fujimura, S.; Nakagawa, Κ.; Ikenaga, Ν.; Suzuki, T. Sekiyu Gakkaishi. 1997, 4. 179. Matsui, N.; Anzai, K.; Akamatsu, N.; Nakagawa, K.; Ikenaga, N.; Suzuki, T. Appl. Catal. A. 1999, 179. 247. Matsui, N.; Nakagawa, K.; Ikenaga, N.; Suzuki, T. J. Catal. 2000, 194. 115. Tsang, S.C.; Claridge, J.B.; Green, M.L.H. Catal. Today. 1996, 23. 3. Sugino, M.; Shimada, H.; Tsuruda, T.; Miura, H.; Ikenaga, N.; Suzuki, T. Appl. Catal. A. 1995, 121. 125. Sakurai, Y.; Suzaki, T.; Nakagawa, K.; Ikenaga, N.; Aota, H.; Suzuki, T. Chem. Lett. 2000, 526.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 15

CO Reforming and Simultaneous CO and Steam Reforming of Methane to Syngas over Co Ni O Supported on Macroporous Silica-Alumina Precoated with MgO 2

2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

x

1

1

1-x

1

2

V. R. Choudhary , A. S. Mamman , B. S. Uphade , and R. E. Babcock 1

Chemical Engineering Division, National Chemical Laboratory, Pune 411 008, India Chemical Engineering Department, University of Arkansas, 3202 Bell Engineering Center, Fayetteville,AR72701

2

Reaction of CO with methane in the absence and presence of steam over Co NiO (x = 0.0-0.5) supported on commercial macroporous silica-alumina catalyst carrier (SA-5205, obtained from M / S Norton Co. U.S.A.) precoated with M g O at different temperatures (700°-900°C) and contact times (0.03-0.3 s at STP) has been investigated. With the increasing Co/Ni ratio of the catalyst, its surface area is decreased and its degree of reduction by H at 900°C (for 1h) is increased. TPR (from 100° to 900°C) of the catalysts indicated that the N i O -CoO-MgO in the catalyst exists as a solid solution. The influence of Co/Ni ratio of the catalyst and feed ratios [CO /CH = 0.8 - 1.2 in the CO reforming and (CO +H O)/CH = 1.2 and CO /H O = 0.4 - 2.0 in the simultaneous CO and steam reforming] on the conversion and selectivity or H /CO product ratio has also been thoroughly investigated. When the Co/Ni ratio is increased, in the CO reforming, the carbon deposition on the catalyst is reduced drastically and the conversion of methane and CO , H selectivity and H /CO ratio is passed through a maximum. The conversion in the simultaneous CO and steam reforming is also passed through a maximum, at the Co/Ni ratio of about 0.17. The supported catalyst with Co/Ni ratio of 0.17 shows high methane conversion activity (methane conversion > 95%) and 100% selectivity for both CO and H with H /CO 2

x

1-x

2

2

4

2

2

2

4

2

2

2

2

2

2

2

2

2

2

224

2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

225 ratio varying from 1.2 - 2.2 in the simultaneous CO and steam reforming processes at a high space velocity. The H /CO ratio can be controlled by manipulating the H O/CO ratio in the feed; it is increased with increasing the H O/CO ratio and vice versa. 2

2

2

2

2

2

Release of large quantities of C 0 in the atmosphere has created a large greenhouse effect causing a global warming. Hence worldwide efforts are being made for conversion of C 0 to useful products. One of the ways of activating C 0 is its reaction with methane to carbonmonoxide and hydrogen, commonly known as C 0 reforming of methane. In the last few years, the research activities on the C 0 reforming of methane have gained a lot of momentum (1). Several studies on this reaction over different catalysts, such as supported Pt group metals (2-10), N i / M g O (6, 11), N i / A l 0 (12), Ni/MgO-CaO (13), N i O CaO (14) and NiO/MgO/SA-5205 (15), have been reported. A rapid coke deposition on the catalyst is, however, a serious problem in the C 0 reforming of methane, particularly when N i - containing catalyst is used (6, 14). C 0 reforming simultaneously with steam reforming and/or oxidative conversion of methane over NiO-CaO (14, 16, 17), NiO-MgO (11) and L a N i 0 (18), Ni/ALPO-5 (19) and NiO/MgO/SA-5205 (15) catalysts have also been reported earlier. The coke formation i n C 0 reforming over NiO-CaO catalyst was reduced drastically by carrying out the C 0 reforming simultaneously with steam reforming (14) or oxidative conversion (17) of methane to syngas. Our earlier study (20) revealed that the addition of cobalt to unsupported Ni-containing catalyst causes a drastic reduction in the carbon formation in the oxidative conversion of methane to syngas. It is, therefore, interesting to study the effect of cobalt addition to Ni-containing catalysts on carbon deposition on them i n the C 0 reforming. Also, in our earlier studies, N i O supported on a low surface area macroporous silica-alumina catalyst carrier precoated with M g O showed high activity and selectivity in the steam reforming, C 0 reforming and simultaneous steam and C 0 reforming of methane to syngas in the presence or absence of 0 (15, 21) .The present work was undertaken for studying the influence of cobalt addition to this catalyst on its activity and selectivity and also on the filamental carbon formation in the C 0 reforming with or without simultaneous steam reforming at different process conditions.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

2

2

2

2

2

2

3

2

2

3

2

2

2

2

2

2

2

Experimental Supported Co Ni!_ O (14.0 ± 0.5 wt%)/MgO/SA-5205 (x = 0.0, 0.05, 0.15, 0.29, and 0.5) catalysts (Table 1) were prepared by depositing mixed nitrates of x

x

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

226 Ni and Co with desired Co/Ni ratio (0.0, 0.053, 0.17, 0.41, or 1.0) from their aqueous solution on 22-30 mesh size particles of commercial catalyst carrier, SA-5205 [sintered low surface area macroporous support, obtained from M/S Norton Co. USA] precoated with MgO, using an incipient wetness impregnating technique, followed by drying and decomposing(or calcining) in air at 900°C for 4h. The catalyst carrier was precoated with MgO by impregnating the carrier with Mg-nitrate, drying and decomposing as above. The support consists mainly of alumina (86.1%) and silica (11.8 wt%) and its surface area, porosity, pore volume and average pore size are 99.95 %), C 0 and (99.99 %) with or without steam. Water was added to the feed using a SAGE syringe pump and a specially designed evaporator. Before carrying out the reaction, the catalyst was heated insitu at 900°C in a flow (50 cm .min" ) of moisture-free nitrogen for 1 h. The catalytic reactions were carried out at different temperatures, gas hourly space velocities (GHSV, measured at 0°C at 1 atm) and relative concentrations of methane, steam, C 0 in the feed. The product gases (after condensation of the water from them at 0°C) were analyzed by an on-line gas chromatography with TCD, using a Spherocarb column and He as a carrier gas. The conversion/selectivity data were collected after a reaction period of 30 min. For the C 0 reforming reaction, a fresh catalyst was used for every experiment. The H and CO selectivities reported in this paper are based on the methane conversion.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

2

3

1

1

3

2

_1

l

2

2

2

2

3

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

1

227

Results Catalyst Characterization Results showing the influence of Co/Ni ratio in the catalyst on surface area are given i n Table 1. The surface area is decreased appreciably with increasing the Co/Ni ratio in the catalyst. Figure 1 shows curves for the T P R (by H ) from 100° to 900°C of the catalyst with different Co/Ni ratios (0.05 - 1.0). The TPR of the catalyst without cobalt (NiO/MgO/SA-5205) is given earlier (15). The TPR curves are quite similar to that observed for the catalyst without cobalt (15, 21) and also for a N i O - M g O solid solution (22, 23), except for a small hump between 400° and 650°C (Fig. 1). The hump is more pronounced for the catalyst with higher Co/Ni ratios and it is shifted towards the lower temperature side with increasing the Co/Ni ratio. The peak maximum temperature for all the TPR curves is 900°C. However, this is not a true peak maximum temperature as the maximum temperature chosen for the T P R itself is 900°C. Hence, true maximum temperature is expected to be above 900°C. The degree of reduction of Co Ni!_ O from the catalyst is increased with increasing its Co/Ni ratio (Table 1). The degree of reduction was estimated from the knowledge of the concentration of CoO and N i O in the catalyst and the amount of H consumed in the TPR, and assuming the reaction stoichiometry (CoO or N i O + H ->Co° or Ni° + H 0 ) .

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

2

x

x

2

2

2

T A B L E I. Surface Area and Degree of Reduction (in the TPR) of the Co Ni O/MgO/SA-5205 (x = 0.0-0.5) Catalysts x

χ

0.0 0.05 0.15 0.29 0.50

lx

Co/Ni Mol ratio

Degree of reduction (%)

0.0 0.053 0.17 0.41 1.0

2.0 1.9 1.6 1.3 1.1

34.5 49.7 61.7 65.1 72.5

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Figure l.TPR of the catalysts with different Co/Ni ratios.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

229 C 0 Reforming of Methane 2

In order to find the influence of Co/Ni ratio on the extent of filamental carbon formation on the catalyst, particularly in the C 0 reforming of methane at different temperatures, a pressure drop across the catalyst bed after a reaction period of 1.0 h was measured. Results showing the influence of Co/Ni ratio on the pressure drop across the catalyst bed due tofilamentalcarbon formation in the C 0 reforming at different temperatures (700°-900°C) are presented in Fig. 2. At all the temperatures the pressure drop and consequently the extent of filamental carbon formation on the catalyst is decreased exponentially with increasing the Co/Ni ratio. It is also strongly influenced by the reaction temperature; it is increased with increasing the temperature. The pressure drop in all the cases was reduced to its original value after oxidative removal of the carbon from the catalyst in a flow of nitrogen-air mixture (25% air) at 500°600°C. Results showing the influence of Co/Ni ratio of the catalyst on the conversion, H selectivity, and H /CO product ratio in the C 0 reforming of methane to syngas for a low contact time (high space velocity) are presented in Fig. 3. At all the temperatures, the conversion, selectivity and H /CO ratio are passed through a maximum at the Co/Ni ratio of about 0.17. This indicates that at this optimum Co/Ni ratio, the catalyst shows the best performance in the C 0 reforming. The performance of the catalyst with the optimum Co/Ni ratio (0.17) in the C 0 reforming of methane at different process conditions is presented in Fig. 4. Figure 4 also shows the influence of temperature, space velocity and CH4/C0 feed ratio on the conversion of both methane and C 0 , H selectivity and H /CO product ratio in the process. Following general observations can be made from the results (Figs. 3 and 4): 2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

2

2

2

2

2

2

2

2

2

2

2

• •

The conversion of C 0 is higher than that of CH . The H selectivity and H /CO ratio (except at 900°C for the intermediate Co/Ni ratios) are less than 100% and 1.0, respectively. 2

2

4

2

These observations indicate that, along with the C 0 reforming of methane (CH +C0 ^2CO+2H ), a reverse water gas shift reaction, (C0 +H ->CO+ H 0) also occurs simultaneously, depending upon the process conditions. This side reaction is favored at lower temperature, higher space velocity and higher C0 /CH4 ratio; the H /CO ratio is increased with increasing temperature but decreased with increasing space velocity and C0 /CH feed ratio (Fig. 4). 2

4

2

2

2

2

2

2

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

230

Figure

2.Influence ofCo/Ni

the catalyst bed in the C0

2

temperatures

(C0 /CH 2

4

ratio in the catalyst on the pressure drop across reforming of methane over the catalysts at different

feed ratio - LOandGHSV

--3.9xl(f

3

1

cm g'

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

!

h~ ).

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

231

Figure 3.1nfluence of Co/Ni ratio in the catalyst on its performance in the C0 reforming of methane at different temperatures (C0 /CH feed ratio = 1.0 and GHSV =3.9xl(f cm g~ h~').

2

2

3

4

l

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

232

-(b)

)-(α)

-(c)

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

/Ο

1

1

L.

1 -?Γ\Γ\

ι

ι

,

ι

(

ar\r\

TEMPERATURE {%)

1

1

>

^

-

^s,

I

1

—4—

:

1

ι

ι

I

ι

.

1-0 1-5 C 0 / C H RATIO

GHSV χ KJ*cn£g!h"

1

2

4

Figure ^Influence of a) reaction temperature (CH/CH, = 1.0 and GHSV = 3.9 χ 10" cm g~ h~') b) space velocity (at 800 'C and C0 /CH = 1.0) and c) C0 /CH ratio (at 80(fC and GHSV =3.9x10' cm g~> h~>) on the conversion CO selectivity and H /CO product ratio in the C0 reforming of methane over the catalyst with Co/Ni ratio ofO. 17. 3

!

e

2

4

3

2

4

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

233 Simultaneous C 0 and Steam Reforming of Methane 2

Figure 5 shows the influence of Co/Ni ratio on the catalyst performance (conversion of methane, C 0 and water and H / C O product ratio) in the simultaneous C 0 reforming ( C H + C 0 - » 2 C O + 2 H ) and steam reforming (CH + H 0 - » C O + 3 H ) . The catalyst performance is strongly influenced by the Co/Ni ratio, similar to that was observed for the C 0 reforming of methane. In this case also, the catalyst shows highest activity when its Co/Ni ratio is 0.17; the conversion of all the reactants is passed through a maximum for the catalyst with (Co/Ni ratio = 0.17). The H / C O ratio is, however, passed through a minimum at this Co/Ni ratio (Fig. 5). Results showing the performance of the catalyst with its optimum Co/Ni ratio in the simultaneous C 0 and steam reforming process at different process conditions are presented in Figs. 6 and 7. Results in Fig. 6 show a strong influence of space velocity on the conversion of all the reactants but almost no effect on the H / C O ratio. The H / C O ratio is, however, decreased markedly with increasing the C 0 / H 0 feed ratio for a constant (C0 +H 0)/CH4 feed ratio (Fig. 7). The pressure drop across the catalyst bed for all the catalysts at the different process conditions was negligibly small, indicating a little or no formation of the filamental carbon on the catalyst during the simultaneous C 0 and steam reforming process. 2

2

2

4

2

4

2

2

2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

2

2

2

2

2

2

2

2

2

2

Discussion The above results reveal that the surface properties, filamental carbon formation in the C 0 reforming of methane and catalytic activity in the C 0 and/or steam reforming of methane are strongly influenced by increasing the Co/Ni ratio of the catalyst. The decrease in the surface area with increasing the Co/Ni ratio (Table 1) indicates that the addition of cobalt causes sintering or crystal growth of the catalyst. Since, the surface area of the support is very low (0.01 m g" ), the observed surface area of the catalyst is mainly because of the active catalytic components (i.e. oxides of nickel and Co and MgO) deposited on the support. The dependence of the trend of the T P R curves (Fig. 1) and degree of catalyst reduction (Table 1) on the Co/Ni ratio shows that the rate of catalyst reduction is enhanced because of the addition of cobalt to the catalyst. The observed hump at low temperature may be due to a faster reduction of that of cobalt oxide, which has not formed a solid solution in the MgO. Both NiO and CoO can form a complete solid solution in M g O (22). However, since the ionic 2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

1

234

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

α) 8 0 0 ° C

b)850°C

0-2

0-6

K> Co/Ni

0

0-2

0-6

RATIO

Figure 5.Influence of Co/Ni ratio in the catalyst on its performance in the simultaneous C0 and steam reforming of methane at 800° and 850°C(CO /CH feed ratio = 0.55, H 0/CE feed ratio = 0.57 and GHSV = 5. J χ Iff cm" g' h' ) 2

2

4

2

1

4

1

1

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

235

b) 8 5 0 ° C

β

α)800 0

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

2

20

cr

οο

—ο

1-6

°

ο

Μ

0

U

Ο"

I

1-2 ι

ι

ι

ι

•

ι

;

100

* v \ A

H

4

xJNNi /- co

90

Ε ζ ο en cr

C

2

80

Η 0-^ 2

LU > Ζ Ο

υ

70 I

0

I

20

I

I

I

40

I

60 GHSV χ

60

20

40 1

l Ô ^ c m V h

60

)

F/gwre 6. Influence of GHSV on the conversion and H /C0 product ratio in the simultaneous steam and C0 reforming of methane over the catalyst with Co/Ni ratio of 0.17 at 800° and 850°C (C0 /CH and H 0/CH feed ratios = 0.57 and 0.57, respectively). 2

2

2

4

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

236

VO

1-5

C 0 / H 0 RATIO 2

2

Figure 7.Influence of CO2/H2O feed ratio on the conversion and H /CO product ratio in the simultaneous C0 and steam reforming of methane over the catalyst with Co/Ni ratio of 0.17 at 850°C f(C0 + H 0)/CH feed ratio = 1.2 and GHSV =2.3 χ 10 cm g~ h ]. 2

2

2

4

3

J

2

4

1

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

237 2+

2 +

radius of C o (0.72°A) is larger than that of N i (0.69°A) (the ionic radius of M g is 0.66°A), the formation of solid solution of CoO in M g O is difficult as compared to that of N i O in MgO. The higher pressure drop across the reactor for the catalyst containing nickel alone (i.e. without cobalt) is resulted form the filamental (whisker) carbon on the catalyst (24,25) leading to plugging of the catalyst bed. The carbon formation can be due to the Boudouard reaction ( 2 C O - » C + C 0 ) , methane decomposition (CH ->C+2H ) and C O reduction (CO+H ->C+H 0), occurring simultaneously. The observed large decrease in the filamental carbon formation with increasing the Co/Ni ratio of the catalyst (Fig. 2) is attributed mostly to its reduced activity for the above carbon forming reactions. Such beneficial effect was also observed earlier in case of the oxidative conversion of methane to syngas over N i and Co containing Y b 0 , Z r 0 and T h 0 catalysts (20). Work is in progress for understanding this beneficial effect of the cobalt addition. 2 +

2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

4

2

2

2

3

2

2

2

The nickel and cobalt oxides in the catalyst are reduced to their metallic form in the initial short reaction period (< 30 min.), first by methane and then by hydrogen produced in the reforming of methane over the partially reduced catalyst. In the C 0 reforming and simultaneous C 0 and steam reforming processes, the catalyst shows best performance at the optimum Co/Ni ratio (0.17) (Figs. 3 and 5). Thus, the addition of cobalt to the catalyst at the optimum concentration has also another beneficial effect in these processes; not only it reduces drastically the filamental coke formation but it also enhances the catalytic activity and increases the selectivity, particularly for the C 0 reforming reaction. It may be noted that, in the simultaneous C 0 and steam reforming processes, there is no formation of any side product and hence, as long as the conversion of both C 0 and water is positive, the selectivity for H and C O in the conversion of methane to syngas is always 100%. The results reveal that the catalyst with optimum Co/Ni ratio of 0.17 is a highly promising catalyst for the C 0 reforming process and also for the simultaneous C 0 and steam reforming process, showing high activity for the conversion of methane to C O and H at a low contact time (high space velocity) with little or no carbon formation on the catalyst. Among the two processes, the simultaneous C 0 and steam reforming process is of great practical importance because of its following interesting features: 2

2

2

2

2

2

2

2

2

2

•

Unlike the C 0 reforming process, the carbon formation on the catalyst is much less and the H selectivity (based on methane conversion) is always 100% as long as the conversion of both C 0 and water is > zero. 2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

238 •

Unlike the steam reforming process, there is no formation of undesired product such as C 0 (which has high green house effect), and hence the CO selectivity (based on methane conversion) is 100%. The H /CO product ratio can be varied between 1.0 and 3.0 by simply manipulating the C 0 / H 0 feed ratio. 2

•

2

2

2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

Conclusions From this investigation, following important conclusions have been drawn: 1.

The surface and reduction properties, filamental carbon formation in the C 0 reforming of methane, and methane-to-syngas conversion activity (in the steam and/or C 0 reforming of methane) over the Co Nii_ O/MgO/SA5205 catalyst (x = 0-0.5) are strongly influenced by the Co/Ni ratio in the catalyst. 2

2

2.

x

x

The addition of cobalt to the supported nickel catalyst has following beneficial effects: •

Formation of filamental carbon in the C 0 reforming of methane is drastically reduced with increasing the Co/Ni ratio in the catalyst. The methane conversion activity of the catalyst in both the C 0 reforming and simultaneous C 0 and steam reforming reactions is increased; it is maximum for the Co/Ni ratio of 0.17. 2

•

2

2

3.

The supported Co-Ni containing catalyst with an optimum Co/Ni ratio of 0.17 is a highly promising catalyst for the conversion of methane to syngas, particularly by the simultaneous C 0 and steam reforming process; the NiO, CoO and MgO in the catalyst exist as a solid solution. 2

References 1.

Edwards, J. H; Maitra, A.M. The chemistry of methane reforming with CO and its current and potential applications. Fuel Processing Technol. 1995, 42, 269-289. Richardson, J.T. and Patipatyadar, S.A. Carbon dioxide reforming of methane with supported rhodium. Appl.Catal. 1990, 61, 293-309. Solymosi, F., Kutsan, Gy. and Erdohelyi, A. Catalytic reaction of CH with CO over alumina-supported Pt metals. Catal. Lett. 1991, 11, 149-156. 2

2. 3.

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

239 4.

Perera, J. H . S. Q., Couves, J. W., Sankar, G. and Thomas, J. M. The catalytic activity of Ru and Ir supported on Eu O for the reaction, CO +CH ->2H +2CO; a viable solar-thermal energy system. Catal. Lett. 1991, 11, 219-226. Ashcroft, A . T., Cheetham, A . K . , Green, M. L. H . and Vernon, P. D . F. Partial oxidation of methane to synthesis gas using carbon dioxide. Nature 1991, 352, 225-226. Rostrup-Neilsen, J. R. and Bak-Hansen, J. H . CO reforming of methane over transition metals. 1993, 144, 38-49. Erodohelyi, Α., Csernyi, J. and Solymosi, F. Catalytic activation of CH with CO over alumina-supported Pt metals. J. Catal. 1993, 25, 265-270. Nakamura, J., Aiakwa, Α., Sato, K . and Uchijima, T. Role of support in the reforming of CH with CO over Rh catalysts. Catal. Lett. 1994, 25, 265-270. Vernon, P. D. F., Green, M. L. H . , Cheetham, A . K . and Ashcroft, A . T. Partial oxidation of methane to synthesis gas and carbon dioxide as an oxidising agent for methane conversion. Catal. Today 1992, 13, 417-426. Masai, M., Kudo, H . , Miyake, Α., Nishyama, S. and Suruya, S. Methane reforming by carbon dioxide and steam over supported Pd, Pt and Rh catalysts. Stud. Surf. Sci. Catal. 1988, 36, 67-71. Choudhary, V . R. and Mamman, A . S. Oxidative conversion of methane to syngas over noble metal supported MgO, CaO and rare earth oxides. Fuel. 1998, 77, 1477-1488. Chen, Y . and Ren. Conversion of methane and carbon dioxide into synthesis gas over alumina-supported nickel catalysts. Effect nickel­ -alumina interactions. J. Catal. Lett. 1994, 29, 39-48. Yamazaki, O., Nozaki, T., Omata, K . and Fujimoto, K . Reduction of carbon dioxide by methane with N i on MgO-CaO containing catalysts. Chem. Lett. 1992, 1953-1954. Choudhary, V . R. and Rajput, A. M. Simultaneous carbon dioxide and steam reforming of methane to syngas over NiO-CaO catalyst. Ind. Eng. Chem. Res. 1996, 35, 3934-3939. Choudhary, V . R., Uphade, B.S. and Mamman, A . S. Simultaneous steam and CO reforming of methane to syngas over NiO/MgO/SA-5205 in presence and absence of oxygen. Appl. Catal. A: 1998, 168, 33-46. Choudhary, V . R. and Rajput, A . M.and Prabhakar, B . Coupling of exothermic oxidative conversion and endothermic CO and steam reforming of methane to syngas over NiO-CaO catalyst. Angew. Chem. Intl. Ed. Engl. 1994, 33, 2104-2106. Choudhary, V . R. and Rajput, A . M.and Prabhakar, B. Energy efficient methane-to-syngas conversion with low H /CO ratio by simultaneous 2

2

5.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

6. 7.

4

3

2

2

4

2

8.

4

9.

10.

11.

12.

13.

14.

15.

2

2

16.

2

17.

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

240 catalytic reaction of methane with C O and oxygen over NiO-CaO. Catal. Lett. 1995, 32, 391-396. Choudhary, V . R., Uphade, B.S. and Belhekar, A . Oxidative conversion of methane to syngas over L a N i O perovskite with or without simultaneous steam and CO reforming reactions. J. Catal. 1996, 163, 312-318. Choudhary, V . R., Uphade, B. S. and Mamman, A . S. Partial oxidation of methane to syngas with or without simultaneous CO & steam reforming reactions over Ni/AlPO-5. Microporous and Mesoporous Mater. 1998, 23, 61-66. Choudhary, V . R., Rane, V . H . and Rajput, A . M. Beneficial effects of cobalt addition to Ni-catalyst for oxidative conversion of methane to syngas. Appl. Catal. A: 1997, 162, 235-238. Choudhary, V . R., Uphade, B . S. and Mamman, A . S. Oxidative conversion of methane to syngas over nickel supported on low surface area porous catalyst carriers precoated with alkaline and rare earth oxides. J. Catal. 1997, 172, 281-293. Highfield, J. G., Bossi, A . and Stone, F. S. Preparation of catalysts III. Scientific bases for the preparation of heterogeneous catalysts. Proceedings of Third International Symposium, Louvain-la-Neuve, Sept. 6-9, 1982. Eds. Poncelet, G., Grange, P, Jacobs, P. A . Stud. Surf. Sci. Catal. 1983, 16, 181-192. Parmaliana, Α., Arena, Α., Frusteri, N. and Giordano, N. J. Temperature­ -programmed reduction NiO-MgO interactions in magnesia supported N i catalysts and NiO-MgO physical mixtures. Chem. Soc. Faraday Trans. 1990, 86, 2663-2669. Rostrup-Nielsen, J. R. Production of synthesis gas. Catal. Today 1993, 18, 305-324. Tsang, S. C., Claridge, J. B. and Green M . L . H . Recent advances in the conversion of methane to syngas. Catal. Today, 1995, 23, 3-15. 2

18.

3

2

19.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch015

2

20.

21.

22.

23.

24. 25.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 16

Low-Temperature C H Decomposition on High-Surface Area Carbon Supported Co Catalysts Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch016

4

Z.-G. Zhang, K. Haraguchi, and T. Yoshida Hokkaido National Industrial Research Institute, 2-17 Tsukisamu-Higashi, Toyohira-ku, Sapporo 062-8517, Japan

High surface area carbon supported Co catalysts with loading up to 35wt% were prepared for converting CH4 to C2 hydrocarbons with the two-step reaction sequence. The catalysts were characterized by N2 adsorption, XRD analysis, TPR and TPD techniques. The activity of the catalysts in the CH4 decomposition was examined in a temperatures range of 300-450°C, while the subsequent hydrogenation in the second step was conducted at 100°C. It was found that the carbon supported catalysts provided higher Co dispersions than SiO2 supported one, and therefore exhibited higher CH4 decomposition activities. Moreover, the molar ratio of produced H2 to consumed CH4 in the decomposition indicated that CHx formed at lower temperatures tended to contain more hydrogen. The production of C2 in the hydrogenation step provided further evidence for the CHx formation in the decomposition and exhibited potential of the carbon supported catalysts to be used in the two-step CH4 reaction. +

+

© 2002 American Chemical Society

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

241

242

Introduction Increasing attention has been paid recently to the two-step reaction of CH4 to produce higher hydrocarbons (C2 ). In this reaction, CH4 is first activated on suitable metal catalysts at mild temperatures (90% conversion of C H and C 0 and high selectivity to CO and H at atmospheric pressure. However, at high pressure a drastic decrease in conversion, a significant change in selectivity to H and CO and substantial carbon formation were observed with both catalysts. 4

2

3

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

4

2

2

2

Experimental Two series of catalysts, Ni/Na-Y and Ni/Al 0 , were prepared by incipient wet impregnation method using Na-Y zeolite (LZ-Y52, Si0 /Al 0 molar ratio: 5; Surface area: 826 m /g) and γ-Α1 0 support (S-l/KF; 203 m /g), respectively. A weighed amount of nickel nitrate to yield the desired loading is dissolved in distilled water. Then the precursor salt solution was added dropwise to the support while stirring. After the impregnation, the resulting powder was dried in an oven for 16 h at 110°C. The samples were calcined at 450°C for 4 h in an aircirculating oven (air flow rate 100 ml/min). The finished catalysts were sealed in sample bottles and kept in a desccicator before use. The BET surface areas of the supports and catalysts were determined by nitrogen adsorption using a GEMINI system (Micromeritics, USA) at -196°C. H and CO chemisorption measurements of catalysts were carried out on an AutoChem 2910 (Micromeritics, USA) analyzer using pulse chemisorption. Temperature programmed oxidation (TPO) of carbon species on the catalysts after the C 0 reforming of CH* reaction was carried out on a thermogravimetric analyzer (Mettler TG-50). Scanning electron microscopic analysis was conducted on fresh and used catalysts in CO2 reforming of C H to investigate the structure of carbon deposits. DS 130 dual stage SEM (ISI, USA) was used for this purpose. The reaction of C 0 reforming of C H was studied at a temperature of 750°C or 800°C under atmospheric pressure (1 atm) and high pressure (400 psi or 27 atm) with the feed C 0 / C H molar ratio of 1.0. Ultra-high-purity gases were used in all experiments. Mass flow controllers allow exact control of the volumes of C H , C 0 and Ar into the reactor at a space velocity (WHSV) of 30,000 cm^g^.h" . The calcined Ni/Na-Y catalyst sample was mixed with 20 wt% of alumina binder (Catapal B), pressed at 5000 psi for 1 min, and sieved between 0.5 to 1 mm. Ni/Al 0 catalyst was used without binder. About 0.1 g 2

3

2

2

2

3

2

2

3

2

2

4

2

4

2

4

4

2

1

2

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

260 of the catalyst was used in each run. The catalyst sample (0.1 g) was placed in an Inconel 800 H alloy tubular reactor (0.54" OD χ 0.375" i.d. and about 16" in length) for high pressure reaction and quartz reactor with similar dimensions for atmospheric reaction. A quartz wool plug was placed at the base of the reactor to support the reactor packing material. Atop the quartz wool plug a bed of ccA1 0 spheres (2 mm diam) were filled up to the lower l/3 of the reactor on which the catalyst was loaded between two quartz wool plugs. Above the catalyst bed a-A1 0 spheres were filled to form a pre-heat bed. A plug of the quartz wool was placed in the inlet of the reactor to hold the beds in place. The catalyst bed temperature was monitored by using a chromel/alumel thermocouple placed in the center of the catalyst bed and a temperature controller. Simultaneously the furnace temperature was also monitored with the help of a separate temperature indicator. The pressure in the reactor was maintained and monitored by using a backpressure regulator and high-pressure transducer, respectively. Before admitting the reactants the catalysts were pretreated in Ar gas flow (30 ml/min) [without or with H addition at 450°C] by using the following temperature program: 100°C with 15 min hold, 450°C with 75 min hold, and at reaction temperature (750°C or 800°C) with a 15 min hold. [The heat up from 100°C to 450°C, and that from 450 to 750°C or 800°C were completed in about 15 min each]. The Na-Y supported catalyst was pretreated in Ar without H , while the Al 03-supported catalyst was pretreated in the same temperature program but with addition of H flow at 450°C [10 ml/min H plus 30 ml/min Ar] and with H flow stopped at the end of 75 min at 450°C, because such pretreatment procedures give the best catalyst performance for both types based on our previous research (11). The commercial alumina-supported Ni catalysts were pretreated in the same way as the laboratory-prepared A1 0 supported catalyst. After this catalyst pretreatment the reaction run was started by introducing C H and C 0 into the reactor (feed: CH4=10 ml/min, CO =10 ml/min and Ar=30 ml/min). The reactor system was connected to a gas chromatograph (GC) (SRI GC 8610C, Torrance, CA, USA) on line. The analysis of the reaction products was carried out by using a packed silica gel column and thermal conductivity detector (TCD). After the steady state conditions were reached (15-30 min) the reaction products were analyzed by on-line GC every 30 min during a time-onstream (TOS) period of 300 min. H and CO yields were calculated from the number of moles of CO or H produced based on the total number of moles of feed (CH +C0 ). rd

2

3

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

2

3

2

2

2

2

2

2

2

4

2

3

2

2

2

4

2

Results and Discussion Table I shows the physicochemical properties of the Ni catalysts supported on Na-Y and γ-Α1 0 prepared by incipient wetness impregnation method (IWI). The surface area of 8 wt% Ni/Na-Y catalyst is much higher than that of 6.6 wt% Ni/Al 0 catalyst, as expected from the difference between the surface areas of 2

2

3

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

261

2

2

Na-Y support (826 m /g for LZ-Y52) and γ - Α 1 0 support (203 m /g for S1/KF). Measurements of Ni dispersion from both CO and H chemisorption are based on the standard stochiometry of H/Ni or CO/Ni as unity (12). Both H and CO pulse chemisorption measurements indicate that the Ni metal dispersion is better on the zeolite Na-Y than on amorphous γ-Α1 0 . 2

3

2

2

2

3

Table I: Physicochemical properties of Ni/Na-Y and Ni/Al 0 fresh catalysts prepared by IWI method Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

2

2

Ni/Ah0 6.6 185 0.27 0.23

Ni/Na-Y 8.0 476 0.38 0.35

Fresh Catalysts Ni loading (wt%) BET area (m .g -cat.) H/Ni (fresh catalyst) CO/Ni (fresh catalyst) _1

3

3

Figure 1 shows C 0 conversion for Ni/Na-Y catalyst at 750°C at atmospheric pressure (1 atm) and high pressure (400 psi or 27 atm). The 8 wt% Ni/Na-Y prepared by incipient wet impregnation (IWI) is one of the better Ni catalysts among various Ni catalysts that we have prepared using different types of zeolites or amorphous supports and different preparation methods. Table II lists the conversion data corresponding to a time-on-stream (TOS) of 90 min. It can be seen from Figure 1 and Table II that C 0 conversion over 8 wt% Ni/NaY at atmospheric pressure is fairly high, and remains steady at 91% over 6 h. The corresponding C H conversion, 89%, is slightly lower than that of C 0 . It is worth mentioning here that several runs at atmospheric pressure were carried out in an Inconel steel reactor [with 8 wt% Ni/Na-Y (IWI), 6.6 wt% N i / A l 0 (IWI) catalysts] as well as in quartz reactor [with 7.9 wt% Ni/Na-Y (WI), 6.6 wt% Ni/ A 1 0 (WI) catalysts prepared by wet impregnation method (WI)]. The differences in C 0 and C H conversions for similar catalysts with the two different reactors under otherwise identical conditions were generally 90%) of C 0 and C H , whereas conversions drops with increasing pressure. At all pressures C 0 conversion was higher than C H conversion. CO yields decreased with increasing pressure (CO yield: 59.2% at 200 psi to 32% at 4

2

4

2

4

2

4

2

2

2

4

2

4

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

265 800 psi), whereas H yield increased with increasing pressure up to 400 psi and beyond which it decreased (H yield: 27.9% at 200 psi to 25% at 800 psi with a maximum of 40.7% at 400 psi). H /CO ratio remained close to unity at all pressures except at 200 psi where it was 0.47. It is worth mentioning again that at atmospheric pressure maximum C 0 and CH4 conversions and CO and H yields were obtained. Increasing pressure decreases catalytic conversion but increasing temperature increases C H and C02 conversion. Based on these results if we select a moderate pressure (for example 200 or 400 psi) and relatively high temperature (800-85CPC), higher C 0 and C H conversions may be achieved. 2

2

2

2

2

4

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

2

0

5

4

10

Ni

loading

15

20

(wt%)

Figure 4. Effect ofNi loading on C0 and CH conversion over Ni/Na-Y (IWI) catalyst in C0 reforming ofCH at 27 atm and 750°C (TOS: 90 min). 2

2

4

4

Figure 4 shows the effect of Ni loading in Ni/Na-Y (IWI) catalysts on C H and C 0 conversion in C 0 reforming of C H at 27 atm and 750°C. The catalytic activity strongly depends on Ni loading, and maximum C 0 and C H conversion was obtained at 8wt% Ni loading. Both C 0 and C H conversions increased with increasing Ni loading up to 8 wt% Ni, beyond that the conversions decreased with further increasing Ni loading. It can be seen from Figure 4 that C 0 conversion is higher than CH* conversion for all Ni loadings. CO and H yield were also increased with increased Ni loading giving an optimum at 8 wt% Ni loading (CO yield: 42.6%; H yield: 40.7%), whereas 4

2

2

4

2

2

4

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

4

266 H /CO ratio remained around unity (between 0.78 to 0.93 from 1 wt% to 12 wt% Ni and 1.5 at 16 wt% Ni loading). We have found that Ni/Na-Y prepared by incipient wetness impregnation (IWI) is much better than that by ion-exchange (IE) method for C 0 reforming of C H (11). For Ni loaded on Na-Y support by IWI, isolation of Ni particles in a three-dimensional structure of Na-Y zeolite surface with Na ion between Ni may inhibit bridged CO chemisorption and suppresses carbon formation. In fact, the carbon formation on Ni/Na-Y catalyst prepared by IWI method was much less than that on the corresponding catalyst prepared by ion-exchange method, and also less than that formed on Ni/Al 0 catalyst. However, even with Ni/NaY prepared by IWI, the carbon formation problem was exacerbated significantly with increasing pressure from atmospheric pressure to 27 atm, which falls into the range of the common pressure range in industrial syngas manufacture. On the other hand the N i / A l 0 catalyst suffers from severe coking in both atmospheric- and high-pressure runs, but the difference in total amount of carbon formation due to pressure change appears to be smaller than that observed for Ni/Na-Y (IWI). The deactivation of the N i / A l 0 catalyst is more rapid during high-pressure reforming, as can be seenfromFigure 1. 2

2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

4

2

2

3

3

2

3

Table III: Analysis of used Ni/Na-Y and Ni/Al 0 catalysts after C 0 reforming at 750°C at atmospheric pressure 2

Used Catalysts H/Ni (Ni dispersionfromH2) CO/Ni (Ni dispersionfromCO) Carbon deposit (wt%)

8 wt% Ni/Na-Y (IWI) 0.32 0.3 27.5 [66.31*

3

2

6.6wt%Ni/Al 0 (IWI) 2

3

0.08 0.05 82.3 [84.51*

* Values outside brackets are after A-P reforming (1 atm) at 750 °C for 300 min, and those in brackets are after H-P reforming (27 atm) at 750 °C for 300 min.

We also analyzed the used Ni catalysts after the CO2 reforming. Table III lists the metal dispersion datafromH and CO chemisorption, and the amount of coke determined by temperature programmed oxidation (TPO). Both H and CO chemisorption data indicate that the metal dispersion was better maintained on Ni supported on Na-Y prepared by IWI. The metal dispersion data for used Ni/Al 0 catalyst was much lower after 300 min TOS as compared to the fresh catalyst (Table I). This may be a result of both the change in catalytic metal species (metal sintering) and the surface carbon deposition. Despite the known uncertainties in the stochiometric ratio of CO/Ni and the possible occurrence of CO disproportionation (12), the results obtained by CO chemisorption are in good agreement with H chemisorption. This may suggest the dominance of CO 2

2

2

3

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

267 chemisorption in the form of linearly-bound CO (CO/Ni rface=l) in the case of CO chemisorption. Temperature programmed oxidation (TPO) was carried out for selected catalysts before and after the reforming reaction. The results obtained after reaction at 1 atm and 27 atm at 750°C are presented in Figure 5. It can be seen from Figure 5 that there was severe carbon formation on N1/AI2O3 catalyst both at atmospheric pressure (82.3% in terms of weight loss from TPO) and high pressure (84.5 % in terms of weight lossfromTPO). Extent of carbon formation is less on Ni/Na-Y catalyst at high pressure (66.3%) compared to N i / A l 0 catalyst. Ni/Na-Y gives 27.3% carbon formation at atmospheric pressure reaction. SU

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

2

3

Temperature ( ° C ) Figure 5. Temperature programmed oxidation profiles from TGA of used 8 wt% Ni/Na-Y, 6.6 wt% Ni/Al 0 catalysts after C0 reforming ofCH at 1 atm (AP) and 27 atm (HP) and 750°C (used after TOS of300 min). 2

3

2

4

Figure 6 shows the SEM photographs of Ni catalysts before and after use in C 0 reforming at 27 atm at 750°C. The catalysts examined were 8 wt% Ni/Na-Y (IWI) (fresh catalyst and used catalyst: a, b), Ni/Na-Y (ion-exchange, IE) (fresh and used catalyst: c, d) Ni/Al 0 (IWI) (fresh catalyst and used catalyst: e, f). It 2

2

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

268

Figure 6. SEM photographs offreshand used catalysts after C0 reforming of CH at 27 atm and 750 °C (after TOS of 300 mm): 8 wt% Ni/Na-Y (IWI; a: fresh b: used), 7.9 wt% Ni/Na-Y (IE; c:fresh;d: used), 6.6 wt% Ni/Al 0 (IWI; fresh; f: used). 2

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

3

269 is worth to mention here that the SEM photographs shown for each catalyst are representative, since several photographs of the same catalysts samples were considered in this regard. In the case of fresh catalysts crystallites of Na-Y can be seen with finely dispersed Ni-zeolite particles. In the used Ni/Na-Y (IWI) catalyst, carbon deposits in the form of carbonfilamentscan be clearly seen. On the other hand SEM photograph of Ni/Na-Y (IE) used catalyst shows dense carbon filamentous growth compared to Ni/Na-Y (IWI) catalyst. Inspite of severe carbon filament growth, Ni/Na-Y catalyst exhibited reasonably good activity and selectivity in CO2 reforming of C H till the end of the 300-min run. This may be due to the unique three-dimensional structure of Na-Y zeolite, wherein the Ni particles located in zeolite cage can be protected from carbon deposition, whereas the Ni particles located on the outer surface of Na-Y zeolite suffer from carbon deposition. It appears from Figure 6 that apart from filamentous carbon some amorphous carbon can also be found on the catalysts surface. On used N1/AI2O3 catalyst, the amount of amorphous carbon is more than thefilamentouscarbon (16). Different types of carbon have been observed on "coked" catalysts ranging from amorphous or microcrystalline species to highly aromatic/graphitic carbons and to carbides [17]. These usually result from polymerization and/or cracking of hydrocarbons. In the presence of C H in its C 0 reforming, C H disproportionate as well as CH4-CH4 coupling may give rise to filamentous carbon growth on Ni particles [18]. The reactant adsorption effects may also greatly influence the carbon formation on metal catalysts [19]. The origin of carbon formation during CO2 reforming of CH4 may be from carbon monoxide disproportionation [ 2 CO = C + C 0 ] and/or methane decomposition [ C H = C +2 H ], depending on the reaction conditions, the type of surface metal species, and the structure of catalyst surface.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

4

4

2

2

4

2

2CO + * -> C0 + C*

(1)

C H + * -> C* + 2H

(2)

2

4

2

Carbon monoxide disproportionation is exothermic whereas methane decomposition is endothermic. The former is favored by higher pressure and lower temperature, while the latter is favored by higher temperature and lower pressure. The reactant adsorption effects may also influence the carbon formation on metal catalysts (13). The movement of catalyst particles is a known phenomenon and surface diffusion, temperature driven dissolutionprecipitation have been reported to be responsible for carbon formation (14). In surface diffusion, surface migration of adsorbed species is an established mechanism. Thus adsorption of a hydrocarbon on a clean metal surface may lead to the diffusion of hydrocarbon or its fragments across the surface to the carbon-metal interface, where decomposition occurs to give growth of the

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

4

270 filament. From studies of Boudouard reaction, it was observed that nucleation of carbon occurred on specific steps and kinks of the nickel surface (15). Alternatively, adsorption of carbon may be followed by decomposition and by diffusion of carbon to the growing filament (15). At high-pressure conditions, the overall carbon concentration (from reactants) is high compared with that at atmospheric pressure, which could increase the rate of methane and carbon dioxide decomposition on surface. The computational analysis of the thermodynamics indicated that the equilibrium conversions of C H and C 0 decrease significantly with increasing reaction pressure (21). The data in Table 1 show that more carbon deposits were formed during high-pressure CO2 reforming compared to atmospheric-pressure reactions. The carbon formation is enhanced at high pressure according to our computational thermodynamic analysis (21). Apartfromcarbon formation, CO2 and C H conversions are governed by equilibrium limitations. The calculated equilibrium conversions of CO2 at 750°C under 1 atm are around 92% and 85%, respectively, for the cases without and with carbon formation incorporated in computational analysis (21). The calculated equilibrium conversions of C 0 at 750°C under 27 atm are around 56% and 53%, respectively, for the cases without and with carbon formation incorporated in computational analysis (21). Some recent reports have indicated that Ni/MgO (22,23) or Ni/CaO catalysts (24) as well as Ni catalysts supported on MgO- or CaO-precoated AI2O3 (25) are more resistant to coke formation during C 0 reforming of C H at atmospheric pressure. On the other hand, commercial Ni-based steam-methane reforming (SMR) catalysts generally have alkali earth promoters such as MgO and CaO; and in some cases with additional alkali promoters such as K; many commercial SMR catalysts contain Ni supported on modified alumina or calcium aluminate or magnesium aluminate (26, 27). Such catalysts have been used in industrial SMR processes at high pressures, and thus it would be interesting to examine such catalysts for C 0 reforming of C H at high pressure. We have obtained several SMR catalysts from Haldor-Topsoe A/S, United Catalyst Co. and ICI Katalco. They are commercially available samples (27) of Ni-based SMR catalysts containing MgO or CaO that were supplied together with a nominal range of composition. Table IV shows the results of performance tests with some commcercial Ni-based catalysts for CO2 reforming at 800°C at 27 atm with C 0 / C H feed ratio of 1.0 and WHSV of 30,000 cm . g V . We also tested our laboratory-prepared catalysts [8 wt% Ni/Na-Y and 6.6 wt% N1/AI2O3] under the same conditions using the same reactor, and the results are shown in Table IV for comparison. Among the three commercial catalysts shown in Table IV, the Ni catalyst R-67 containing MgO and the Ni catalyst G-91 containing CaO appear to be more active than the Ni catalyst CI 1-9-02 for C 0 reforming of CH4 at 800°C under 27 atm. The laboratory-prepared catalysts give similar range of CO2 and C H conversion at 800 °C under 27 atm when compared to the two commercial catalysts R-67 and G-91. The comparison of high-pressure and atmosphericpressure tests over the laboratory catalysts again demonstrate that increasing

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

4

2

4

2

2

2

2

4

4

4

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

2

4

2

3

2

3

27 atm-800°C 55.7 49.4 48.1 40.8 0.85

2

2

2

3

4

3

3

27 atm-800°C 31.9 36.8 25.5 25.8 1.01

2

2

N i O (10-30%), A 1 0 (70-90%)

3

N i O (15-25%), CaO (5-15%), C a A 1 0 (515%), A 1 0 (55-60%), K O(l-10%) 27 atm-800°C 53.3 59.8 58.0 45.2 0.78

2

N i O (15-20%), M g O (20-25%), A 1 0 A 1 0 (5560%)

3

Nominal range of composition data supplied by vendors Press-Temp C 0 conv (%) C H conv (%) C O yield (%) H yield (%) H /CO

2

Commercial Cat United Catalyst C I 1-0-02 Ni/Al 0

Commercial Cat United Catalyst G-91 Ni/Al 0 -CaO

Commercial Cat Haldor-Topsoe R-67 Ni/Al 0 -MgO

Catalyst Type Catalyst ID

3

3

27 atm-800°C 43.9 34.9 46.3 42.0 0.91

2

2

Laboratory Cat PSU/ACEL, IWI 6.6 wt% Ni/Al 0 Laboratoryprep'd, IWI on S - l / K F g - A l 0 support 6.6 wt% N i

2

27 atm-800°C 38.6 43.8 40.6 42.9 1.06

Laboratoryprep'd, IWI on L Z - Y 5 2 N a - Y support 8 wt% N i

Laboratory Cat PSU /ACEL, IWI 8 wt% N i / N a - Y

4

2

2

3

3

Laboratory Cat PSU/ACEL IWI 6.6 wt% Ni/Al 0 Laboratoryprep'd, IWI on S - l / K F g-Al 0 support 6.6 wt% N i 1 atm-800°C 93.1 89.4 89.9 69.4 0.68

1 atm-800°C 92.3 95.6 73.9 67.4 0.91

Laboratoryprep'd, IWI on L Z - Y 5 2 N a - Y support 8 wt% N i

Laboratory Cat PSU/ACEL IWI 8 wt% N i / N a - Y

T a b l e IV: C o m m e r c i a l S M R and L a b o r a t o r y N i Catalysts Tested for C 0 R e f o r m i n g of C H at 8 0 0 ° C under H i g h Pressure (27 atm) (Corresponding T O S of 90 m i n for the D a t a Reported).

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

272 pressure can decrease C 0 and CH4 conversions dramatically, cause product yields to decrease and change H /CO ratios. The performance of CI 1-9-02 was poor in terms of lower C 0 conversion compared to other commercial or laboratory catalysts. Our analysis of used catalysts by TPO, however, revealed severe carbon formation on all these commercial SMR catalysts under high-pressure C 0 reforming conditions employed, with all the used catalysts after 300 min TOS containing over 60 wt% carbon deposits. It should be mentioned that these commercial catalysts were designed and optimized for steam reforming of natural gas and hydrocarbons, and our results in Table IV only reflect on their performance for C 0 reforming of C H under the conditions employed here in a small laboratory reactor. Besides the effects of temperature and pressure, the level of carbon formation will be very dependent upon the ratios of steam/C0 /CH . 2

2

2

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

2

2

4

2

4

Concluding remarks Remarkable differences in C 0 reforming of C H at atmospheric pressure and high pressure were observed. The differences in catalyst conversions at 1 atm and 27 atm are in large part due to thermodynamic limitations. Both Ni/Na-Y and Ni/Al 03 were reasonably active at atmospheric pressure with high CO2 and C H conversion (>89%). The conversions over the same catalysts show a drastic decrease at high pressure (27 atm), but the change in pressure had different impacts on C 0 and CH* conversions, and varied effects on H /CO ratios of the syngas products. Carbon formation on some catalysts that display a superior performance at atmospheric pressure may be exacerbated at high pressure. Low conversions of C 0 and C H and low yields of CO and H were obtained from C 0 reforming of C H at high pressure of 27 atm with both laboratory and commercial Ni-based steam reforming catalysts, and all the catalysts suffered significant carbon formation within 300 min of TOS at high pressure. Increasing reaction temperature increased catalytic conversion whereas increasing pressure has an opposite effect. On Ni/Na-Y, significant effects of Ni loadings on catalyst activity were observed; among the catalysts 2ith 1-16 wt% Ni loading, the 8 wt% Ni loading on Na-Y by IWI method gives optimum catalyst activity. TPO characterization of carbon deposits in used catalysts showed that more severe carbon formation occurs at high pressure in the case of Ni/Na-Y (IWI), whereas the use of N i / A l 0 results in severe carbon formation both at 2

4

2

4

2

2

2

2

4

4

2

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

273

atmospheric as well as high pressure C 0 reforming of C H . SEM investigation of used catalysts reveals that there are remarkable effects of type of support (NaY or AI2O3) used and catalyst preparation method on the structure of carbon deposits. 2

4

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

Acknowledgment We are pleased to acknowledge the start-up funding provided to CS for this work by the Energy Institute (for purchasing reactor parts) and the Department of Materials Science and Engineering (for purchasing mini-GCs) of the Pennsylvania State University. We thank Air Products & Chemicals Inc. and members of APCI/PSU SAMCOM for funding the work on high-pressure C 0 reforming of C H . We are grateful to Prof. Harold H. Schobert for his encouragement for this work, and to Mr. Ron M. Copenhaver and Dr. Jian-Ping Shen for assistance in setting up the flow reactor and on-line GC. We would like to acknowledge the Haldor-Topsoe A/S, United Catalyst Co. and ICI Katalco for graciously providing their commercial samples of steam hydrocarbon reforming catalysts with nominal compositional data. 2

4

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Gunardson, H.H.; Abrardo, J.M. Hydrocarbon Processing, April 1999, p 87. Armor, J.N. Appl. Catal A: 1 9 9 9 , 176, 159. Rostrup-Nielsen, J.H.; Hansen, J.H.B. J. Catal. 1 9 9 3 , 144, 38. Ross, J.R.H.; van Keulen, A.N.J.; Hegarty, M.E.S.; Seshan, K. Catal. Today. 1 9 9 6 , 30, 193. Wang, S.; Lu, G. Q. CHEMTECH, 1 9 9 9 , 29,37. Xu, X.; J. A. Moulijn. Energy & Fuels, 1 9 9 6 , 10, 305. Bradford, M.C.J.; Vannice, M.A. J.Catal. 1 9 9 8 , 173, 157. Chang, J. -S.; Park, S. -E.; Chon, H. Appl. Catal. A. 1 9 9 6 , 145, 111. Bhat, R. N.; Sachtler, W. M. H. Appl. Catal. A. 1 9 9 7 , 150, 279. Ruckenstein, E.; Hu, Y.H. Ind.Eng.Chem. Res. 1 9 9 8 , 37, 1744. Song, C.; Murata, S.; Srinivas, S.T.; Sun, L.; Scaroni, A.W. Am. Chem. Soc. Div. Petrol. Chem. Prepr., 1 9 9 9 , 44, 160. Zhang, Z.; Verykios, X.E. Appl. Catal. 1 9 9 6 , 138, 109. Nishiyama, Y.; Tamai, Y. Carbon, 1 9 7 6 , 14, 13. Rostrup-Nielsen, J.; Trimm, D.L. J. Catal. 1 9 9 7 , 48, 155. Gadalla, A.M.; Bower, Β. Chem. Eng. Sci., 1 9 8 8 , 35, 3049. Srinivas, S.T.; Song, C.; Pan, W.; Sun, L. Am. Chem. Soc. Div. Petrol. Chem. Prepr., 2 0 0 0 , 45, 348. Lobo, L.S.; Trimm, D.L. J. Catal., 1 9 7 3 , 29, 15. Bernardo,C.;Trimm, D.L. Carbon, 1 9 7 6 , 14, 225.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by OHIO STATE UNIV LIBRARIES on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch017

274 19. Nishiyama, Y.; Tamai, Y. Carbon, 1916,14, 13. 20. Rostrup-Nielsen, J. Stud. Surf. Sci. Catal., 1994, 81, 25. 21. Pan, W.; Srinivas, T.S.; Song, C. 16 th International Pittsburgh Coal Conference, Pittsburgh, USA, October 11-15, 1999, Paper No. 26-2. CD-ROMISBN 1-890977-16-0 1999. 22. Tomishige, K.; Chen, Y.; Li, X.; Yokoyama, K.; Sone, Y.; Yamazaki, O.; Fujimoto, K. Stud. Surf. Sci. Catal. 1998, 114, 375. 23. Ruckenstein, E.; Hu, Y.H. Catal. Lett. 1998, 51, 183. 24. Cheng Z.X.; Wu Q.L.; Li J.L.; Zhu, Q. Catal. Today, 1996, 30, 147. 25. Choudhary, V.R.; Uphade B.S.; Mamman A.S. Catal. Lett. 1995, 32, 387. 26. Gunardson, H. Industrial Gases in Petrochemical Processing, Marcel Dekker, New York, 1998,p46. 27. OGJ Special. Oil Gas J., Sept 27, 1999,p63.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 18

A Comparative Study on CH -CO Reforming over Ni/SiO -MgO Catalyst Using Fluidizedand Fixed-Bed Reactors 4

2

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

2

A. Effendi, Z.-G. Zhang, and T. Yoshida Resource and Energy Division, Hokkaido National Industrial Research Institute, 2-17 Tsukisamu-Higashi, Toyohira-ku, Sapporo 062-8517, Japan (email: [email protected])

CH -CO reforming to syngas was studied by using micro-fluidized-and fixed-bed reactors over Ni/SiO -MgO at 700 °C, 1 atm with CO /CH =1. The catalytic activity in the fluidized-bedreforming showed higher CH and CO conversions and larger H /CO ratio compared to those in the fixed-bed reforming. An efficient contact between reactants and catalyst in the fluidization was suggested to be partly responsible for the enhancements. Moreover, the fluidization minimized a temperature gradient of the catalyst bed and a concentration gradient of the reactant gases. Although only a limited amount of carbon, which consisted of amorphous and crystallized forms, was deposited on spent catalysts, a relatively higher XPS intensity of crystallized carbon was observed on the fixed-bed catalyst than that on the fluidized-bed one. This suggested a faster deactivation of the catalyst in the fixed-bed reactor. In addition, an uneven coke deposit was observed alongfixed-bedpositions. 4

2

2

2

4

4

2

2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

275

276

Introduction Catalytic C 0 reforming of C H (Eq. 1 ) has been renewed as an attractive route for direct production of synthesis gas. Under the given conditions, however, a competitive coke formation occurs as results of C H decomposition (Eq. 2) and CO disproportionation (Eq. 3), causing a catalyst deactivation. In addition, a reverse water-gas-shift (RWGS) reaction (Eq. 4) occurs simultaneously during the reforming. 2

4

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

4

CH

4

+C0

2

CH

4

->

C + 2H

2CO ~>

C0

2

+H

->2CO + 2 H

2

Δ Η Γ

Δ Η

2

C+C0

2

Γ

A H

2

r

->CO+H 0 2

2 9 8

2 9 8

2 9 8

Δ Η Γ

= + 247.9kJ/mol (1)

= + 75.0 kJ/mol (2)

=-172.0 kJ/mol (3)

= +41.0 kJ/mol (4)

2 9 8

Ni-based material has been considered as potential catalysts for industrial applications, and their modifications to high resistant coke ones by adding basic oxides such as MgO have been widely investigated. A loading of 5 wt % N i exhibited a high activity and much less magnitude of coke deactivation (1-4). Beside the massive coke deposition, performing C H - C 0 reaction by using a fixed-bed reactor has limitations due to its high endothermicity. Poor heattransfer can cause a temperature gradient across a fixed-bed, creating a cool-spot. To minimize these difficulties, the use of a fluidized-bed reactor has been approached (1,5,6). A fluidized reforming allowed catalyst-circulations, resulting in almost isothermal reactor operation, less concentration gradient of reactant gases and lower deactivation of catalysts (1,5,6). Lab-scale experiments showed that thermodynamic equilibrium could be achieved (5). On the other hand, a mechanical loss due to catalyst attrition during the fluidization might take place. In the present work, the catalytic C H - C 0 reforming has been examined by using a micro fluidized-bed reactor over Ni/Si0 -MgO at 700 °C, 1 atm with C0 /CH =1. For a comparison, fixed-bed experiments on the reforming were carried out under the same conditions. In both reactions, catalytic parameters such as C 0 and C H conversions as well as H /CO ratio were investigated. Coke deposits on spent catalysts were also characterized. 4

4

2

2

2

2

4

2

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

277

Experimental

Catalyst Preparation and Characterization A catalyst with a loading of 4.5 wt % NiO on Si0 -MgO (Nikki Chemicals Co. Inc., 75.6 wt % S i 0 , 21.3 wt % MgO and particle size of 45-88 μπί) was prepared by a multiple impregnation using an aqueous solution of nickel nitrate precursor. The support containing the active metal was dried at 70 °C between impregnation steps, then at 110 °C overnight and finally calcined in air at 550 °C for 4 hrs. Bulk compositions were determined by using ICP-AES (Shimadzu GV-1000P), whereas the surface by XPS analysis ( V G E S C A L A B 2201 X L ) . Surface area (70.8 m /g) and average pore volume (16.2 cm /g) were measured by N adsorption according to the BET method (Belsorb 28 apparatus, Bel Japan, Inc.). Prior to the analysis, samples were degassed at 200°C for 4 hrs. Phase analysis of powder catalysts was done by X R D (Rigaku Ltd.) using C u K a radiation (35 k V , 20 mA). Samples for T E M (JEM2010, JEOL Ltd.) were powdered and dispersed in ethanol to prepare suspensions, then applied to sample grids for this analysis. Crystallites of NiO with average particle sizes of 100-130 Â were observed from latter analyses. 2

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

2

2

3

2

Apparatus and Conditions of Reforming A quartz micro-reactor (id of 9 mm) was applied for fluidized- and fixedbed experiments. A thermocouple inserted inside a thermowell was used for measuring the catalyst-bed temperature and the furnace temperature was controlled by another thermocouple placed outside the wall of the reactor. A power supply (Chino SU KP3000) equipped to a tubular gold-coated glass furnace was employed. This furnace allowed an even heat distribution through the reactor-wall. Gases (CH , C 0 , and Ar) were purified by passage through appropriate adsorbents, and delivered to the reactor through mass flow controllers (Tokyo Keiso Co. Ltd.). Both reactions were performed under same conditions, namely at a furnace temperature of 700 °C, 1 atm with a C0 /CH =1, and equal contact time (88 ml/min, STP, 150 mg catalyst, so W/F= 0.64 g.h/mol). In the case of the fluidized-bed experiments, the feed was supplied up-flow through a porous quartz disc. A down-flow gas stream was employed for fixedbed experiments. Initially, the catalyst was purged with Ar (40 ml/min, STP) up to 700 °C, and then C 0 / C H gases were switched to the reactor. Gases were analyzed on-line by a G C (Yanaco G3800 GC) using a 2m χ 3mm id activated carbon column at isothermally 110 °C and 50 Nml/min Ar carrier. Reactants and products (CO, H 0 and H ) from the reactor effluent were passed through a condenser and anhydrous Mg(C10 ) sorbent to trap residual moisture. Amount of condensed water as well as the total flow rate of effluent-gas was quantified. 4

2

2

2

2

4

2

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

4

278 A catalytic test using the support alone resulted in below 0.1 % of CO and H yields, showing a negligible activity.

2

Coke Determination Coke amount on spent catalysts after reaction tests was determined by using a TG analyzer (Shimadzu DT40). Approximately 15 mg sample was pre-dried up to 110°C under 10% 0 /Ar (30 ml/min) for 10 min and a weight-loss profile was recordedfromthis temperature to 825°C with a heating rate of 10°C/min. Then, the profile was corrected with a baseline measuredfroma blank test. XPS surface analysis with C l s spectra was used to examine the nature of carbonaceous species. The AlKcc was used as an energy source (1486.6 eV), operated at 10 kV, 15 mA and mainly an energy pass of 30 eV. As a reference line, the Au4f spectrum with a binding energy (BE) of 84.0 eV was used for all measurements of pelletized samples. Moreover, changes in surface compositions of metal atoms were also examined.

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

2

7/2

Results and Discussion

Conditions for Fluidized- and Fixed-Bed Experiments Choice of the present experimental conditions was based on results given in Fig. 1. The figure shows conversions of C0 , C H and H /CO ratio as a function of space velocity (W/F) using a fixed-bed reactor. Both conversions increased with the increase in W/F, then unchanged beyond 1.30 g.h/mol. Similar tendency was also obtained for the H /CO ratio. The suppression of their conversions in a range beyond 1.30 g.h/mol suggested that the reforming was strongly influenced by the diffusion of the reactant gases at the interphase of catalyst particles. Thus, W/F=0.64 g.h/mol was chosen for both experiments, in which the diffusion effect was considered to be less significant and the requirement for a minimum fluidization was met. Consequently, the fluidizedbed of 150 mg catalyst achieved a minimum height of H «4 mm, velocity of U m f ^ ml/min, STP, as visually measured and the u/u of > 1.5 was employed. 2

4

2

2

mf

mf

Catalytic Activity using Fluidized- and Fixed-Bed Reactors All catalytic activities on CH -C0 reforming were tested at 700 °C, with C 0 / C H = 1.0, 1 atm for 30 hrs time-on-stream (tos) and the results are shown in Fig. 2 and Table I. Catalytic activity obtained from the fluidized-bed 4

2

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

279

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

280 experiment showed superiority on conversions as well as H /CO ratio compared to thatfromthe fixed-bed one. The conversions of C H and C 0 were higher by 16 and 19 %, whereas the H /CO indicated 0.69 and 0.54 for the fluidized- and fixed-bed reactions, respectively. For all cases, C 0 conversions were always higher than the C H ones, moreover, the resulted H /CO and C 0 / C H ratios were differed from the stoichiometric values due to the occurrence of accompanied reactions (Eq. 2-4). The higher H /CO ratio obtained from the fluidized-bed reaction was attributed to relatively less selectivity towards the RWGS reaction, as the water yield (3.4 mol %) was lower than that from the fixed-bed one (3.8 mol %). To elucidate the catalytic activity difference using both type reactors, particular parameters such as catalyst-bed temperatures, pressure-drops over the reactors, and coke deposits were further discussed. 2

4

2

2

2

4

2

2

4

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

2

Effects of Catalyst-bed Temperature Due to the highly endothermic reforming, the catalyst-bed temperature gradient and its effect on the catalytic activity were a major importance to investigate. Under Ar alone at 700 °C, no temperature difference was observed between the catalyst-bed and furnace for both cases. However, when the supply of a C 0 / C H mixture started, the bed temperature was gradually decreased and remained constant after 1 hr tos. 2

4

Table I. Results of C H - C 0 reforming over Ni/Si0 -lVlgO catalyst 4

Bed reactor

Tb

2

2

C0 /CH 2

4

H /CO 2

(°Q

Fluidized Fixed Fluidized at 684 °C Fixed-fluidized 2

1

695 684 684 696

Conversions (%) CH co 37.7 52.7 21.5 34.0 51.9 37.5 23.2 37.9 4

0.84 0.77 0.78 0.81

0.69 0.54 0.68 0.63

2

Coke wt% 1.5 1.6 nd nd ;

2

not determined for 1.5 hrs in fixed-bed, then fluidized mode for 4.0 hrs tos

Although an axial temperature gradient over catalyst-beds was not determined, measuring temperatures at the center of the catalyst-bed might reflect a thermal reaction behavior in fluidized- and fixed-bed reactors. As shown in Fig. 3, the fluidized-bed temperature increased to 695 °C, then attained a steady state, whereas the fixed-bed temperature showed a slight increase, and remained unchanged at 684 °C. As the temperature difference was related to the thermal behavior, this might also reflect a measure of the overall change in enthalpy (7). According to this approach, the bed temperatures showed different reaction-heats, so a lower catalyst temperature of the fixed-bed indicated a larger reaction-heat absorbed for this catalytic reforming. The fixed-bed disfavored the thermal efficiency due to an inferior heat-transfer, as consequence lowering the

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

281 70

60

co V ° ° O o

2

0

O o O

o

o

0

o

0

o

0

o

o

0

o

o

0

o

0

50

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

CH (Λ C

o

0

o

o

/

fluidized bed

4

40

Ο Φ >

ο ο

30

-

1

/

fixed bed

'! CH

4

20

10 500

1000

2000

1500

time on stream (min.)

Figure. 2. Conversions of CH and C0 over Ni/Si0 -MgO resulted from the reforming usingfixedandfluidized-bedreactors 4

2

2

700 fluidized-bed :

X

X X

X

X X

X

X

X X

X

X

X

X

X

X

X

X X

X

X

X X

X

X

X

X

X

690 fixed-bed

ο J3

680

670 500

-+-

4-

1000

1500

2000

time on stream (min.)

Figure. 3. Catalyst-bed temperature profiles of fluidized- andfixedreactors during CH -C0 reforming over Ni/SiO MgO 4

2

r

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

282 overall bed temperature. In the case of the fluidized-bed, the catalyst circulation in gases led to less heat-transfer resistance, and the absorbed-heat for the endothermic reforming was favored. Moreover, an even-heating bed was possibly achieved assuring an isothermal region, inhibiting a temperature gradient. The above temperature difference of 11 °C might represent their variance in the catalytic activities. To examine the influence of the bed-temperatures, the fluidized-bed reforming at 684 °C was also conducted. Table I shows that lowering the catalyst-bed temperature to 684°C, which was the same bedtemperature as that of the fixed-bed applied, did not decrease significantly the catalytic activity, and the conversion levels remained close to those resulted at 695 °C. Hence, this temperature variation did not cause the difference in the catalytic activity between the fluidized- and fixed-bed reactions.

Profiles of Pressure-drop Pressure-drop over both reactors was monitored for 30 hrs tos in which its alteration might reflect coke amount. At both initial activity tests, the pressuredrops increased slightly, and then remained significantly unchanged for 30 hrs tos. This was possibly associated to the limited amount of coke formed, «1.5 wt %, analyzed by the TG as shown in Table I.

Deposited Coke Evidence of a Rapid Catalyst Deactivation in Fixed-bed. The lower catalytic activity in the fixed-bed might be related to a rapid coke formation at initial stages. To understand coke poisoned-levels on the catalyst deactivation in the fixed-bed, a short run of a catalytic test was carried out firstly in a fixed-bed for 1.5 hrs, then immediately switched to a fluidized-mode for 4 hrs. As shown in Fig. 4, the catalytic activity was lowered in the fixed-bed reaction, and then was recovered when switched to a fluidized-mode. However, the final C H and C 0 conversions appeared to be slightly higher than the result obtained from the fixed-bed mode, and much smaller than their conversions obtained by the fluidized-bed experiment alone (see Table I). It suggests that the formation of coke at the initial stage of the fixed-bed reaction gave decisive effects on lowering the catalytic activity. 4

2

Characterization on the nature of coke. Since the coke amount of both spent catalysts indicated insignificant difference, the XPS analysis was applied to distinguish the nature of the surface carbonaceous species. Figure 5 shows the C l s spectra for the fresh and spent catalysts. C l s line of the fresh catalyst

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

283

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

70 (a) fixed 60 (b) fluidized

_

50

(b)

(a)

CO

c ο

S2 ο >

C0 40

-

2

o o o .

37.9 %

ο

Ο

c ο

Θ

ο

Ο

30

CH *

_

20

•

4

• ·

23.2%

•

10 150

450

300

time on stream (min.) Figure. 4. Conversions of CH and COj over Ni/Si0 -MgO obtained from the fluidized reforming after experiencing thefixed-bedmode 4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

284

294

292

290

2S8

286

284

282

Binding Energy - eV

Figure. 5. XPS Cl s spectra offresh Ni/SiOr-MgO fromfluidized-andfixed-bedreforming

and spent catalysts resulted

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

285 demonstrated a main peak with a B E of 290 eV assigned to carbonate, C 0 " species possibly due to the exposure to air containing C 0 . Whereas, C l s lines of both spent catalysts presented peak-maxima at 287.0 and 284.7 eV, indicating the presence of amorphous (-C-CO-) and crystallized (-C-C-) structures, respectively. In the case of the fluidized-bed sample, the crystallized one was relatively a minor species. On the other hand, the average fixed-bed sample showed that the crystallized carbon became a major species and the amorphous form was less pronounced. Table II shows the relative intensity of the crystallized to amorphous forms. According to the temperature-programmed oxidation (TPO) studies (24,6,9,10), it is well known that crystallized form is a less reactive carbon and usually oxidized at higher temperatures, whereas the highly reactive amorphous carbon is readily oxidized at lower temperatures. This kind of reactive species would be formed at initial stages of catalytic runs, then could be removed by gasifying with C 0 . At the same time, a possible remaining carbon would be gradually converted to a crystallized form. It became a major species on N i catalyst after 1 hr (2-5). 3

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

2

2

Table II. XPS relative intensities of crystallized to amorphous forms and surface elemental compositions (atomic %) of Ni/Si0 -MgO catalysts 2

Catalyst

Lc-c-fi-c-co- 0

Ni

Si

Mg

C

Fresh

-

63.8

2.5

19.5

4.5

9.7

Spent (fluidized-bed)

0.67

58.6

0.7

16.4

3.0

21.3

Spent (fixed-bed)

1.93

55.4

1.1

16.0

2.8

24.7

In the catalyst fluidization, a concentration gradient of the reactant gases and an accumulation of less reactive carbon leading to the formation of crystallized carbon could be prevented. During the C H - C 0 reforming, the formation of carbon by C H cracking and its gasification with C 0 were in balance and these occurred simultaneously. When the carbon was formed, it would be immediately gasified by C 0 , so this process inhibited a conversion of less reactive to crystallized forms. Table II shows that on both spent catalysts, the surface concentrations of metals and oxygen decreased as carbon was accumulated. As more crystallized species was observed on the fixed-bed spent catalyst, which suggested to be a primary cause for the catalyst deactivation due to a poisoned encapsulation of Ni particles (3,4,9,10). Therefore, the catalyst deactivation was relatively faster in the fixed-bed reaction. 4

4

2

2

2

Carbonaceous species along thefixed-bed.As the color intensity of the spent catalyst resulted from the fixed reactor increased from dark gray to black with bed-positions, this might reflect its coke distribution. Accordingly, a separate experiment on a longer fixed-bed (10 mm) was carried out by using a

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

286

-C-C-

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

284.7

-J 294

I

!

I

L

I

292

290

288

286

284

L_ 282

Binding E n e r g y - e V

Figure. 6. XPS Cl s spectra of spent catalysts sampled at top and bottom of the fixed-bed

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

287 smaller reactor (id of 4 mm) under the same condition. Figure 6 demonstrates C l s spectra of the top and bottom samples recovered after 5 hrs tos. The carbonaceous species from the top sample was dominated by a species with a B E of 287.0 eV, whereas the crystallized carbon positioned at 284.7 eV was a significant species from the bottom sample. Hence, various carbonaceous species was distributed as a function of bed positions. It was previously suggested (2-4,9,10) that the formation of amorphous species was associated with C H decomposition. In the bottom of the fixed-bed reactor, the C H concentration was higher than that of C 0 , as a result, the C 0 / C H ratio was less than unity. The concentration gradient of C 0 was possibly caused by the progress of the reforming with fixed-bed positions. At the bottom of the catalyst bed, the C 0 concentration would be minimum, leading to the suppression of the carbon gasification with C 0 . Thus, the carbon concentration increased continually with fixed-bed positions and was relatively higher on catalysts in the bottom of the bed. Moreover, since the presence of C 0 was insufficient to remove all carbon formed, CO disproportionate (Eq.3) would possibly take place. 4

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

4

2

2

4

2

2

2

2

Conclusions A comparative study on the C H - C 0 reforming using two different reactor types leads to the following conclusions. Conducting of the C H - C 0 reforming in a fluidized-bed reactor increased the performance of the catalytic activity and H /CO ratio, compared to that in the fixed-bed one. Amorphous and crystallized forms were observed on the spent catalysts in the fluidized- and fixed-bed reactions. The presence of the crystallized form was a major carbon species in the fixed-bed sample, as the result, the catalyst deactivation in the fixed-bed reforming was relatively faster. In addition, a concentration gradient of the reactant gases over the fixed-bed resulted in uneven carbon distribution. 4

2

4

2

2

Acknowledgements We acknowledge the financial support by the Japan International Science and Technology Exchange Corporation under the Science and Technology Fellowship Program. We also thank N . Imai for TPR, T G measurements, and Mr. O. Nishimura (HNIR1, Sapporo) for XPS analysis.

References 1. 2.

Effendi, Α., Z-G. Zhang, M. Sahibzada, T. Yoshida, Am. Chem. Soc. Div. Petrol. Chem. Prepr., 45(135-137) (2000) Schmitz, A . D. and T. Yoshida, "Greenhouse Gas Control Technologies", Pergamon (1999) 355

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

288 3. 4. 5.

Wang, S. and G. Q. Lu, Ind. Eng. Chem. Res., 38 ( 1 9 9 9 ) , 2615-2625 Lu, G. Q. and S. Wang, Chem. Tech., January ( 1 9 9 9 ) 37-43 Mlecko, L., S. Malcus and T. Wurzel, Ind. Eng. Chem. Res., 36 ( 1 9 9 7 ) 4459-4465 6. Olbye, O., L. Mlecko and T. Wurzel, Ind. Eng. Chem. Res., 36 ( 1 9 9 7 ) 5180-5188 7. Gadalla, A. M. and B. Bower, Chem. Eng. Sci. 42(1988), 3049-3062 8. Kunii, D. and O. Levenspiel, Fluidization Engineering 2 edition, Butterworth-Heinemann ( 1 9 9 1 ) 9. Himeno, Y., K. Tomoshige, K. Fujimoto, Sekiyu Gakkashi, 42(4) ( 1 9 9 9 ) , 252-257 10. Wang, S., G. Q. Lu, G. J. Millar, Energy and Fuels, 10 ( 1 9 9 6 ) , 896-904

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch018

nd

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 19

Effects of Pressure on C O Reforming of C H over Rh/Na-Y and Rh/Al O Catalysts 2

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

2

4

3

Srinivas T. Srimat and Chunshan Song* Clean Fuels & Catalysis Program, The Energy Institute, and Department of Energy and Geo-Environmental Engineering, Pennsylvania State University, 209 Academic Projects Building, University Park, PA 16802 *Corresponding author: fax: 814-865-3248; email: [email protected]

Laboratory-prepared Rh/Na-Y and Rh/Al O catalysts were examined in CO reforming of CH at atmospheric pressure (1 atm) and high pressure (27 atm). It was found that the use of higher pressure substantially decreased CO and CH conversion and increased catalyst deactivation during CO reforming of CH , compared to runs at 1 atm for both Rh/Na-Y and Rh/Al O catalysts. Deactivation was related to carbon formation, and the type of support also affects both the amount and structure of carbon deposits. 2

2

3

4

2

4

2

4

2

3

Reforming of methane by carbon dioxide yields synthesis gas with a low H /CO ratio close to unity, which is suitable for adjusting H /CO ratio for certain synthetic applications such as oxo synthesis and Fischer-Tropsch synthesis (13). This reaction has received increasing attention in the past decade (2-14). 2

2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

289

290 Many catalysts tested for C 0 reforming of C H are based on Ni supported systems (4-6) that are similar to steam-reforming catalysts. Noble metals have been reported to be more resistant to coking than Ni at 1 atm (7-10) for reactions at atmospheric pressure. On the other hand, the industrial application of this reaction would require high-pressure conditions at 20 atm or higher (2,3). Under high-pressure conditions catalysts may be more susceptible to carbon formation (1). Hence a study was carried out in our laboratory to examine the behavior of Rh catalysts supported on zeolites and on amorphous materials for high-pressure CO 2 reforming of C H . This paper reports the results obtained on Rh catalysts for C O 2 reforming of C H at 1 atm and 27 atm. Pressure significantly affects the conversion over Rh catalysts and selectivity to H and CO. At similar reaction conditions, Rh supported on Na-Y and A l 0 3 exhibit comparable activity and selectivity to Ni catalysts supported on Na-Y and AI2O3. 2

4

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

4

4

2

2

Experimental Various catalysts with Rh supported on Na-Y zeolite (LZ-Y52, Si0 /Al 03 molar ratio: 5; Surface area: 826 m /g) and γ - Α 1 0 (S-l/KF; 203 m /g) were prepared by incipient wetness impregnation (IWI) method. In a typical preparation, a weighted amount of rhodium chloride to yield desired loading (1 wt% Rh) was dissolved in distilled water. Then the precursor salt solution was added drop-wise to Na-Y or γ - Α 1 0 support while stirring. After the impregnation, the resulting powder was dried in oven for 16 h at 110°C. Then the samples were calcined at 450°C for 4 h in an air-circulating oven (flow rate 100 ml/min). For comparison purpose, different rhodium precursor and different preparation procedure were also employed in some catalyst preparation and reforming experiments (see below). Thefinishedcatalysts were sealed in sample bottles and kept in desccicator before use. BET surface areas of supports and catalysts were determined by nitrogen adsorption using a GEMINI system (Micromeritics, USA) at -196°C. Rh dispersion was determined from H and CO chemisorption of catalysts carried out on a Micromeritics 2910 AutoChem (USA) analyzer using a pulse chemisorption method. Carbon deposits on the catalysts after the reaction were determined by temperature programmed oxidation using TGA (Mettler TG-50). Scanning electron microscopy was used on fresh and used catalysts after CO2 reforming of C H to investigate the structure of carbon deposits. DS 130 dual stage SEM (ISI, USA) was used. The reaction of C 0 reforming of C H was studied at a temperature of 750°C or 800°C at atmospheric pressure (1 atm) and at high pressure of 27 atm (400 psi) with a feed C 0 / C H molar ratio of 1.0. Ultra-high-purity gases were used in all experiments. Mass flow controllers allow exact control of the volumes of C H , C 0 and Ar into the reactor at a space velocity (WHSV) of 30,000 cm^g'^h" . The calcined Rh/Na-Y catalyst sample was mixed with 20 wt% of binder alumina (Catapal B), pressed up to 5000 psi for 1 min, and sieved 2

2

2

2

2

3

3

2

4

2

2

4

4

4

2

1

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

291 (between 0.5 to 1 mm), and RI1/AI2O3 catalyst was used without binder. About 0.1 g of the catalyst was used in reaction runs. The catalyst sample (0.1 g) was placed between two quartz wool plugs in an Inconel 800 H alloy tubular reactor (0.54" OD χ 0.375" i.d. and about 16" in length). A quartz wool plug was placed at the base of the reactor to support the reactor packing material. Apart from the catalyst bed the reactor was packed with CC-AI2O3 spheres (2 mm dia). Temperature of the reactor was monitored by using a chromel/alumel thermocouple placed in the center of the catalyst bed and a temperature controller and the furnace temperature was monitored with the help of a separate temperature indicator. The pressure in the reactor was maintained and monitored by using a backpressure regulator and pressure transducer, respectively. The reactor system, the procedure for CO2 reforming and the on-line GC used in this work are the same as described in another paperfromour laboratory (11) was connected to GC (SRI GC 8610C, Torrance, CA, USA) on line. Before admitting the reactants the catalysts were treated in Ar flow (30 ml/min) [without or with the addition of H at 450°C] by using the following temperature program: 100°C, 15 min hold-450°C, 75 min hold-750°C, 15 min hold. The zeolite-supported Rh catalysts prepared by impregnation or ion-exchange were pretreated in Ar without H , while the Al 03-supported Rh catalysts were pretreated with the addition of H (10 ml/min plus 30 ml/min Ar) at 450°C for 75 min. After this catalyst pretreatment the reaction run was started by introducing feed (CH =10 ml/min, CO =10 ml/min and Ar=30 ml/min) into the reactor. After the steady state conditions were reached (15-30 min) the reaction products were analyzed by on-line GC every 30 min during a time-on-stream (TOS) period of 300 min. H and CO yields were calculated based on number of moles of CO or H produced to total number of moles of feed (CH +C0 ). 2

2

2

2

4

2

2

2

4

2

Results and Discussion Table I shows the physico-chemical properties of the 1 wt% Rh/Na-Y and 1 wt% RJ1/AI2O3 catalysts prepared by IWI using rhodium chloride precursor. Figure 1 shows the C 0 conversion for atmospheric-pressure and high-pressure runs over Rh/Na-Y and Rh/Al 0 catalysts. The activity data for the Rh catalysts at TOS of 90 min are presented in Table II for comparison. Rh/Na-Y gives 90.4 % C O 2 conversion at 1 atm whereas the CO2 conversion drops to 52.3% at 27 atm. Similarly, RI1/AI2O3 gives 91.3% CO2 conversion at 1 atm while it decreased to 45% at 27 atm. Thus high CO2 conversions were achieved with both catalysts at atmospheric pressure, but with increasing pressure both the conversion and product yields dropped significantly. Similar trends of differences in C H conversion between runs at 1 atm and 27 atm were observed, as can be seen in Table II. CO and H yields for Rh/Na-Y and RI1/AI2O3 are shown in a bar graph (Figure 2). Both H and CO yields are lower from reforming at higher pressure. However, increasing pressure decreased more of H formation over Rh catalysts, 2

2

3

4

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

292 and consequently this resulted in a lower H /CO ratios at high pressures with both Rh/Na-Y and Rh/Al 0 catalysts. 2

2

3

Table I: Physicochemical properties of 1 wt% Rh/Na-Y and 1 wt% catalysts prepared by IWI using rhodium chloride precursor supported on Na-Y and A1 0

RI1/AI2O3

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

2

Catalyst Precursor Fresh Catalyst Metal loading (wt%) BET area (m .g -cat.) H/M (fresh catalyst) CO/M (fresh catalyst) Used Catalyst H/M (used catalyst) CO/M (used catalyst) Carbon deposit (wt%)* _1

Rh chloride Rh/Na-Y 1.0 628 0.66 0.55

3

Rh chloride Rh/Al 0 1.0 177 2

0.42 0.4 5.3 Γ7.41*

3

0.5 \\AY

N O T E : * Values outside brackets are after A - P reforming (1 atm; 750°C; TOS:300 min); values inside brackets are after H-P reforming (27 atm; 750°C; TOS:300 min).

The decrease in the activity between atmospheric-pressure (1 atm) and highpressure (27 atm) operations can be attributed to several factors including equilibrium limitations of the reaction and carbon formation, and possibly some other issues associated with pressure change. We have calculated equilibrium conversions for C 0 and CH4 at 1 atm and 27 atm in the presence and absence of carbon formation (12). The calculated equilibrium conversions for C 0 at 750 °C under 1 atm are, 92% and 85% respectively, for cases without and with carbon formation incorporated in computational analysis. At the same reaction temperature under 27 atm, equilibrium conversions for C 0 are 56% and 53% respectively, for cases without and with carbon formation incorporated in computational analysis (12). The calculated equilibrium conversions of C H at 750°C under 1 and 27 atm incorporating the carbon formation are around 95% and 68%, respectively. The calculated equilibrium conversions of C H at 750°C under 1 and 27 atm without incorporating the carbon formation are around 88% and 34%, respectively (12). It can be seenfromequilibrium conversions that in the case of C 0 , decrease in conversion is small with considering carbon formation and without considering carbon formation at both 1 atm and 27 atm, whereas higher decrease in equilibrium C H conversion can be seen calculated 2

2

2

4

4

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

293

Table II: Conversion and product yields for C 0 reforming at 750°C over 1 wt% Rh/Na-Y and 1 wt% Rh/Al 0 catalysts prepared by IWI using rhodium chloride precursor (TOS: 90 min) 2

2

Rh precursor

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

Catalyst Pressure-Temp C 0 conv.(%) CH4 conv. (%) CO yield (%) H yield (%) H /CO ratio 2

2

2

Rh Chloride, IWI 1 wt% Rh/Na-Y 1 atm750°C 90.4 89.1 78.6 60.9 0.77

3

Rh Chloride, IWI 1 wt% Rh/Na-Y 27 atm750°C 52.3 24.2 25.7 17.9 0.69

Rh Chloride, IWI 1 wt% Rh/AbOj 1 atm750°C 91.3 89.1 82.1 60.9 0.74

Rh Chloride, IWI 1 wt% RI1/AI2O3

27 atm750°C 45.0 21.5 51.6 31.7 0.61

100 90

c ο

"3>5 _ Φ

> c ο ο o υ

-I

80 70

Rh/Na-Y (AP) Rh/Na-Y (HP) - ± — RIVAI203 (AP) Rh/AI203 (HP)

-I

60

w

50

-I

40 30

Figure

I

I

50

100

I

I

150 200 Time (min)

I

I

250

300

350

1. C0 conversion for 1 wt% Rh/Na-Y and 1 wt% Rh/Ah 0 at 1 atm (AP) and 27 atm (HP) at 750 °C. (Rh chloride was usedfor IWI). 2

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

294 under similar conditions. This shows that carbon formation resultsfromC H decomposition and is in agreement with previous literature reports (5-9, 13,14). Thus our experimental data is in agreement with computational thermodynamic calculations showing that the reaction is governed by equilibrium limitations and influences catalyst activity at atmospheric and high pressure apartfromcarbon formation. The TOS profiles in Figure 1 illustrate that the Rh catalysts display a stable performance against deactivation at the atmospheric pressure, but begin to deactivate within the first few hours at high pressure of 27 atm. The deactivation is more rapid with RJ1/AI2O3 catalyst than with Rh/Na-Y catalyst during highpressure CO 2 reforming. Relative to the runs at atmospheric pressure, the more pronounced decrease in catalyst activity with time on stream at 27 atm can be attributed mainly to enhanced carbon formation. On Na-Y, Rh is dispersed on the outer surface of the zeolite as well as in its inner pores. Carbon formation that takes place on the outer surface of the catalyst can block the active sites.

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

4

CO-AP

CO-HP

H2-ÂP

H2-HP

Figure 2. CO and H yields for I wt% Rh/Na-Y and 1 wt% Rh/Al 0 at I atm (AP) and 27 atm (HP) at 750 °C (Rh chloride was usedfor IWI) (TOS: 90 min). 2

2

3

We also analyzed the 1 wt% Rh/Na-Y (prepared using RhCl ) by H and CO pulse chemisorption before and after the CO2 reforming at 750°C (after TOS 3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

295 of 300 min). It can be seenfromthe data for used catalyst in Table I that after 300 min reaction at 750°C, Rh dispersion in Rh/Na-Y decreased by about 10%. The chemisorption data for the 1 wt% RI1/AI2O3 catalyst are not available at this time. On the other hand, we have observed a significant loss of the apparent Rh dispersion in a 2.5 wt% Rh/Al 0 catalyst after the CO2 reforming reaction at 750°C (eg. CO/Rh before reaction=0.27; CO/Rh after reaction at 27 atm=0.09).

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

2

3

Table III: Comparision of C O and H yields in C 0 reforming at 750°C over Rh catalysts (prepared from Rh chloride) and Ni catalysts (prepared from Ni nitrate) supported on Na-Y and A1 0 (TOS: 90 min) 2

2

2

Catalyst Pressure-Temp 1 wt% Rh/Na-Y (IWI) 1 wt%

Hi yield .(%) 1 atm750°C 60.9

H2 yield (%) 27 atm750°C 17.9

3

CO yield (%) 1 atm750°C 78.6

CO yield. (%) 27 atm750°C 25.7

60.9

31.7

82.1

51.6

68.9

40.7

85.6

42.6

66.3

29.6

81.9

57.9

RI1/AI2O3 (IWI)

8wt% Ni/Na-Y (IWI) 6.6 wt% N1/AI2O3 (IWI)

We have also examined the performance of several supported Ni catalysts for atmospheric-pressure (6) and high-pressure (11) C 0 reforming at 750°C as well as 800°C. At high pressure of 27 atm, in general, the Rh catalysts showed no better catalyst performance than the Ni/Na-Y catalysts on similar supports under comparable conditions, both in terms of C H and C 0 conversions and catalyst stability at 27 atm. At atmospheric pressure and 750°C, 90% C H and CO2 conversions were obtained on Rh as well as Ni catalysts supported on Na-Y and A1 0 . At high pressure and 750°C, both CH4 and C 0 conversions drop but remained in similar range for all catalysts. C 0 conversion (%) for various catalysts follows: Ni/Na-Y (46.2), Ni/Al 0 (43), Rh/Na-Y (52.3) and Rh/Al 0 (45); C H conversion (%) for various catalysts follows: Ni/Na-Y (29.1), M/AI2O3 (33.3), Rh/Na-Y (24.2), Rh/Al 0 (21.5). However, both Rh/Na-Y and Rh/Al 0 appear to be much more resistant to carbon formation than Ni/Na-Y and N i / A l 0 catalysts, since the amounts of carbon accumulated on the catalysts during 300-min runs were found to be much smaller on the former (Table I) than on the latter (11), by an order of magnitude. 2

4

2

4

2

3

2

2

2

3

4

2

2

3

3

2

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

3

296 Table III shows CO and H yields on Rh catalysts (preparedfromchloride precursor) and Ni catalysts (prepared from nitrate precursor) at 750°C and atmospheric pressure (1 atm) as well as at high pressure (27 atm). It can be seen from Table III that both H and CO yields are similar on Rh and Ni catalysts supported on Na-Yand A1 0 at atmospheric pressure (1 atm) and 750°C. On the other hand Ni/Na-Y gives more H yield than Rh/Na-Y catalyst at high pressure (27 atm) and 750°C. Where as both Rh and Ni supported on A1 0 give similar £t yields at high pressure. Similar trends were observed in CO yields obtained at high pressure. Bhat and Sachtler (7) have reported on the superior performance of 0.93 wt% Rh/Na-Y catalyst prepared by ion exchange when compared to other Rh catalysts on amorphous supports for C 0 reforming at atmospheric pressure. We also prepared a Rh/Na-Y (IE) with 1 wt% Rh by ion-exchange method following the method and procedure reported by Bhat and Sachtler (7) using rhodium amine precursor and Na-Y support [LZ-Y52]. In addition, we also prepared a Rh/Na-Y (IWI) catalyst using the same rhodium amine precursor and Na-Y support but by incipient wetness impregnation method. 2

2

2

3

2

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

2

3

2

Table IV: Conversion and product yields for C 0 reforming at 800°C over 1 wt% Rh/Na-Y (IE) and 1 wt% 1 wt% Rh/Na-Y (IWI) catalysts prepared using rhodium amine complex precursor 2

Rh precursor

Catalyst Pressure-Temp C 0 conv. (%) CH conv. (%) CO yield (%) H yield (%) H /CO 2

4

2

2

Rh Amine Complex (IE) 1 wt% Rh/Na-Y 1 atm800°C 94.0 91.3 77.8 62.2 0.81

Rh Amine Complex (IE) 1 wt% Rh/Na-Y 27 atm800°C 58.5 35.0 42.9 27.1 0.63

Rh Amine Complex (IWI) 1 wt% Rh/Na-Y 1 atm800°C 93.5 90.3 74.1 59.9 0.81

Rh Amine Complex (IWI) 1 wt% Rh/Na-Y 27 atm800°C 53.7 49.4 62.0 31.5 0.51

Table IV shows the results for conversion and product yields for atmospheric-pressure and high-pressure C 0 reforming at 800°C over 1 wt% Rh/Na-Y (IE) and 1 wt% Rh/Na-Y (IWI) catalysts prepared using rhodium amine complex precursor. For the runs under atmospheric pressure, both Rh catalysts give high C 0 and C H conversions, and the catalyst performance seems to be stable. These observations are similar to those reported by Bhat and 2

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

297 Sachtler (7). However, when the pressure is increased to 27 atm, the conversion dropped significantly, and so did the yields of CO and H . Moreover, the 2

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

OC02 conv. (%)-Rh/AI203

0

!

,

,

:

•

,

0

1

2

3

4

5

1 6

Rh loading (wt%) Figure 3. Effect of Rh loading on C0 and CH conversion over Rh/Na-Y (IWI) andRh/AhOi (IWI) catalysts in C0 reforming ofCH at 27 atm and 750 °C (TOS:90 min). 2

4

2

4

catalysts begin to deactivate within 5 hours of TOS. The 1 wt% Rh/Na-Y (IWI) catalyst appears to be better than the 1 wt% Rh/Na-Y (IE) catalyst under high pressure C 0 reforming conditions in terms of higher CO and H yields and similar catalyst stability. We have also studied the effect of Rh metal loading in Rh/Na-Y (IWI) and RI1/AI2O3 (IWI) preparedfromRh chloride precursor on conversion and product yield in C 0 reforming of CH at 27 atm and 750°C. C 0 and CH4 conversion as a function of Rh loading is shown in Figure 3. It can be seen in Figure 3 that C 0 conversion is higher than C H conversion for both Rh/Na-Y and Rh/Al 03 catalysts. In the case of Rh/Na-Y catalyst, C 0 and C H conversions increase with increasing Rh loading up to 2.5 wt%, beyond that the conversions decreased with further increase in metal loading. Rh/Al 0 catalyst also shows a 2

2

2

2

4

2

4

2

2

4

2

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

298 similar trend, where C 0 and C H conversions increase with increasing Rh loading up to about 2.5 wt%, beyond which both conversions decreased. CO and H yields increased with increasing Rh loading (CO yield: 23.4% at 0.1 wt% Rh to 34.9% at 5 wt% Rh; H yield: 14.1% at 0.1 wt% Rh to 23% at 5 wt% Rh), where as H /CO ratio was below unity at all Rh loadings. 2

4

2

2

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

2

Figure 4. SEMphotographs offreshand used catalysts after CO2 reforming of CH at 27 atm and 750 °C (after TOS of300 min): 2.5 wt% Rh/Na-Y (IWI) (a: fresh; b: used), 2.5 wt% Rh/Al 0 (IWI) (c:fresh;d: used). 4

2

3

Figure 4 shows SEM photographs of 2.5 wt% Rh/Na-Y (IWI) (fresh and used catalyst) and 2.5 wt% Rh/Afc0 (IWI) (fresh and used catalyst) after C 0 reforming of C H at 27 atm and 750°C for 300 min TOS. It is very interesting to 3

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

299 note that, on the used 1 wt% Rh/Na-Y (IWI) catalyst (prepared by IWI using rhodium chloride precursor), the texture of carbon deposit is entirely different when compared to those on used Ni catalysts. The carbon in the form of closelyknit coils dominate on the Rh/Na-Y catalyst, and there is also some amorphous carbon andfilamentouscarbon. The absence offilamentousor coiled carbon on Rh/Al 03 is also interesting. It is worth to mention here that the TPO characterization of this catalyst showed very small amount of carbon deposits (1.4%). Temperature programmed oxidation (TPO) studies of used catalysts by TGA showed only a very small amount of carbon deposits on used Rh/Al 0 catalyst in the form of amorphous carbon. For example, TPO profiles of 1 wt%

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

2

2

3

Figure 5. Temperature programmed oxidation profiles from TGA of used 1 wt% Rh/ΝαΎ (IWI) and I wt% Rh/Al 0 (IWI) used catalysts after C0 reforming of CH at 1 atm and 27 atm and 750 °C (after TOS of300 min). 2

3

2

4

Rh/Na-Y (IWI) and 1 wt% Rh/Al 0 (IWI) catalysts (prepared from rhodium chloride precursor) after C 0 reforming of C H at 1 atm and 27 atm and 750°C are shown in Figure 5. It can be seen from Figure 5 that, Rh/Na-Y catalysts results in more carbon formation (7.5% in terms of weight lossfromTGA), 2

2

3

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

300

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

where as RJ1/AI2O3 results in less carbon formation at high pressure (1.4% in terms of weight loss from TGA). On the other hand at atmospheric pressure Rh/Na-Y catalysts results in more carbon formation (5.3% in terms of weight loss from TGA), where as RI1/AI2O3 results in very little carbon formation at high pressure (0.5% in terms of weight lossfromTGA). Various types of carbon have been observed on "coked" catalysts. They range from amorphous or microcrystalline species to highly aromatic/graphitic carbons and to carbides [15], and these usually arise from polymerization and/or cracking of hydrocarbons. In the presence of C H in its CO2 reforming, C H disproportionation as well as CH4-CH4 coupling may give rise to filamentous carbon growth on Ni particles [16]. On the other hand, the reactant adsorption effects may also greatly influence the carbon formation on metal catalysts [17]. The movement of catalyst particles is a known phenomenon and surface diffusion, temperature driven dissolution-precipitation have also been found to be responsible for carbon formation [18]. In surface diffusion, surface migration of adsorbed species is an established mechanism, and it is suggested that carbon fibers may grow behind a nickel crystallite by a similar mechanism [19]. Thus adsorption of a hydrocarbon on a clean metal surface may lead to the diffusion of these species across the surface to the carbon-metal interface, where decomposition occurs to give growth of the filament. From studies of Boudouard reaction, it was observed that nucleation of carbon occurred on specific steps and kinks of the nickel surface [19]. Alternatively, adsorption of carbon may be followed by decomposition and by diffusion of carbon to the growing filament [20]. The carbon deposits on metal can be caused by the decomposition of CO and/or C H or other gaseous reactions [20]: 4

4

4

2CO->C + C 0

2

(a)

C H —> C + 2 H

2

(b)

4

C 0 + 2H ->C+2H 0

(c)

CO + H ->C + H 0

(d)

2

2

2

2

2

These reactions will be favored by metals acting as oxygen receptors. The carbon species originally produced is believed to be atomic carbon and is highly active in nature. These carbon species are also important intermediates in reactions such as methanation or methane steam reforming [21,22]. In fact, steam is a more effective gasifying agent than C 0 . It is possible that carbon and 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

301 surface oxygen recombine to form CO [23]. As a result, carbons may also be a major intermediate in CO2 reforming of CH4, and the produced water in this reaction is expected to react with reactive surface carbon to give H , CO and C 0 . Consequently, in C 0 reforming of CH , hydrogénation of C 0 proceeds through adsorbed species produced by dehydrogenation of C H and the dissociation of C-H bond in CH4 may be a rate limiting step [24]. Similarly, the dissociative adsorption of C 0 on metal surface results from hydrogénation of C 0 . It is believed that the carbon species deposited on Ni surface are important intermediates in the reaction of C 0 with C H . However, relative surface concentrations and the adsorption coverage for different reactant molecules could effect the catalytic properties of Ni and noble metal catalysts in C 0 reforming of CTLu 2

2

2

4

2

4

2

2

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

2

4

2

Concluding remarks The reaction system pressure significantly affect the C 0 and C H conversion over Rh catalysts and CO and H yields as well as H /CO ratios during C 0 reforming of C H . Both C 0 and C H conversion decrease significantly with increasing pressurefromatmospheric pressure to 27 atm. The reaction pressure has different impacts on H formation and CO formation, and thus increasing pressure not only decreases C 0 and C H conversion, but also can change the H /CO ratio of the products. Catalyst stability during C 0 reforming of CH4 can significantly decrease upon increasing pressure from 1 atm to 27 atm for Rh/Na-Y and Rh/Al 03. Some Rh catalysts that showed superior performance at atmospheric pressure can begin to deactivate within hours at high pressure. Significant effects of Rh loadings on catalyst activity were observed for Rh/Na-Y and Rh/Al 0 , and 2.5 wt% seems to be optimum Rh loading to get maximum activity. TPO characterization of carbon deposits in used catalysts showed that, Rh supported on Na-Y and A1 0 show much smaller amounts of carbon deposition after the reaction compared with Ni catalysts. SEM investigation of used catalysts show different effect of type of support (Na-Y or A1 0 ) on the carbon structure. A very unusual form of carbon on used Rh/Na-Y catalyst was observed by SEM, where the carbon deposits in the form of closely-knit coils dominate. 2

2

2

4

2

4

2

4

2

2

4

2

2

2

2

3

2

3

2

3

Acknowledgment We are pleased to acknowledge the start-up funding provided to CS for this work by the Energy Institute (for purchasing reactor parts) and the Department of Materials Science and Engineering (for purchasing mini-GC) of the

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

302 Pennsylvania State University. We thank Air Products & Chemicals Inc., Dr. John Armor of APCI and members of APCI/PSU SAMCOM for funding the work on high-pressure C 0 reforming of CH . We are grateful to Prof. Alan W. Scaroni and Prof. Harold H. Schobert for their encouragement for this work, and to Mr. Ron M . Copenhaver and Dr. Jian-Ping Shen for assistance in setting up the flow reactor and on-line GC. 2

4

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch019

References

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

Rostrup-Nielsen, J. Natural Gas ConversionII,Curry-Hyde, H., Howe, R., Eds.; Elsevier, New York, 1994, p25. Armor, J.N. Appl. Catal A: Gen. 1999, 176, 159. Gunardson, H.H.; Abrardo, J.M., Hydrocarbon Processing, April 1999, p 87. Choudhary, V.R.; Uphade, B.S.; Mamman, A.S.,Appl. Catal. A: Gen. 1998, 168, 33. Wang, S.; Lu, G. Q. CHEMTECH, 1999, 29, 37. Song, C.; Murata, S.; Srinivas, S.T.; Sun, L.; Scaroni, A.W. Am. Chem. Soc. Div. Petrol. Chem. Prepr., 1999, 44, 160. Bhat, R. N.; Sachtler, W. M. H. Appl. Catal. A. Gen. 1997, 150, 279. Bradford, M.C.J.; Vannice, M.A. J. Catal. 1999, 183, 69. Matsui, N.; Anzai, K.; Akamatsu, N.; Nakagawa, K.; Ikenaga, N.; Suzuki, T. Appl. Catal. A: Gen. 1999, 179, 247. Chang, J. -S.; Park, S. -E.; Chon, H. Appl. Catal. A: Gen. 1996, 145, 111. Song, C.; Srinivas, S.T.; Sun, L.; Armor, J.N. Am. Chem. Soc. Div. Petrol. Chem. Prepr., 2000, 45, 143. Pan, W.; Srinivas, T.S.; Song, C. 16 th International Pittsburgh Coal Conference, Pittsburgh, USA, October 11-15, 1999, Paper No. 26-2. CD-ROMISBN 1-890977-16-0 1999. Ross, J.R.H.; van Keulen, A.N.J.; Hegarty, M.E.S.; Seshan, K. Catal. Today. 1996, 50, 193. Xu, X.; Moulijn, J. A. Energy & Fuels, 1996, 10, 305. Lobo, L.S.; Trimm, D.L. J. Catal., 1973, 29, 15. Bernardo, C.; Trimm, D.L. Carbon, 1976, 14, 225. Nishiyama, Y.; Tamai, Y. Carbon, 1976, 14, 13. Rostrup-Nielsen, J.; Trimm, D.L. J. Catal., 1977, 48, 155. Gadalla, A.M.; Bower, B. Chem. Eng. Sci., 1988, 35, 3049. Satterfield, C. Heterogeneous Catalysis in Practice, McGraw-Hill, New York, 1980, p.280. McCarty, J.G.; Wise, H. J. Catal., 1979, 57, 406. Davis, S.M.; Zaera, F.; Somorjai, G. J. Catal., 1982, 77, 439. Bernardo, C.A.; Trimm, D.L. Carbon, 1979, 17, 115. Kim, G.J.; Cho, D.S.; Kim, K.H.; Kim, J.H. Catal. Lett., 1994, 28, 41.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

Chapter 20

Methane Reforming with Carbon Dioxide and Oxygen under Atmospheric and Pressurized Conditions Using Fixed- and Fluidized-Bed Reactors Keiichi Tomishige, Yuichi Matsuo, Mohammad Asadullah, Yusuke Yoshinaga, Yasushi Sekine, and Kaoru Fujimoto Department of Applied Chemistry, School of Engineering, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-8656, Japan

Effect of the reactor in methane reforming with carbon dioxide and oxygen over NiO-MgO catalysts under atmospheric and pressurized conditions was investigated. Under atmospheric pressure, the stable activity was observed using the fluidized bed reactor. However, in the fixed bed reactor, the activity decreased gradually with time on stream. On the other hand, at low temperature and high space velocity, methane conversion decreased rapidly to the level of methane combustion using fluidized bed reactor. Under pressurized condition, the stable production of syngas was possible even at high space velocity since the catalyst is in more reducing atmosphere at higher pressure. The fluidized bed reactor enhanced more effective conversion of methane to syngas than the fixed bed reactor.

Dry reforming of methane (CH4 + C 0 ~> 2CO + 2H ΔΗ=247 kJ/mol) is a suitable to the production of CO-rich syngas which can be utilized to Fischer2

2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

303

304 Tropsch, methanol and dimethyl ether syntheses (7). One of the problems in C 0 reforming of methane is the heat supply because the reaction is highly endothermic. Internal heat supply by the combination of the reforming with the combustion (CH4 + 2Ο2 - » C 0 + 2H 0 AH=-%6\ kJ/mol) is one of the solutions (2-6). Some researches about this kind of methane reforming including the partial oxidation in thefixedbed reactor have been carried out (2, 3). It has been reported that methane combustion is followed by methane reforming and water-gas shift reaction. In that case, significant temperature gradient in the catalyst bed was generated. Groote et. al. simulated the temperature gradient of catalyst bed in partial oxidation of methane using the fixed bed reactor (2). They pointed out that the inlet temperature of the catalyst bed increased up to 1700 Κ though the reactor was controlled at about 1223 K. In addition, at the beginning of the catalyst bed, the catalyst was oxidized and exhibited low activity in the reforming reaction. Dissanayake et.al. have reported that the catalyst bed was divided into three parts in the partial oxidation of methane over Ni/Al 0 using fixed bed reactor (3). Thefirstpart was composed of NiAl 0 , and the second part consisted of NiO/Al 0 . It is found that oxygen reached these two parts. Thus the catalyst in these two parts contributed to combustion of methane. However in the third part, surface nickel was in the metallic state and showed high activity in methane reforming. Several researches on the reforming of methane with internal heat supply using the fluidized bed reactor have been reported (4-6). It has been insisted that the high rates of heat transfer and the stability of the operation were given by the fluidized bed reactor. Combustion proceeded in the front part of the catalyst bed and the reforming occurred in the rear part in both cases of fixed andfluidizedbed. It is thought that both combustion and reforming proceed on one catalyst particle in fluidized bed reactor. Santos et al. have reported the partial oxidation of methane over Ni/MgO and Co/MgO using fluidized bed reactor (4). It has been discussed that Ni and Co change between oxidized and reduced states, and these continuous redox cycles can affect the stability and catalytic behaviors in the long run. Pressurized syngas is more favorable because the synthesis reactions have been carried out under pressurized condition. The problem in C 0 reforming of methane is the carbon deposition, which becomes more serious under higher pressure. Therefore, it is necessary to develop the catalyst with higher resistance to carbon deposition (7). Our research group has developed NiOMgO solid solution catalysts with high resistance to carbon deposition in C 0 reforming of methane (8-16). In this article, we investigated the effect of the internal heat supply in the C 0 reforming of methane over NiO-MgO catalysts under atmospheric and pressurized conditions using the fixed and the fluidized bed reactors. 2

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

2

2

2

2

3

4

2

3

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

305

Experimental Ni Mgi_ O (x=0.03, 0.10) catalysts were prepared by the coprecipitation method from aqueous solution of N i ( C H C O O ) * 4 H 0 (Kanto Chemical Co., inc. >98.0 %) and M g ( N 0 ) · 6 H 0 (Kanto Chemical Co., inc. >99.0 %) using K C 0 (Kanto Chemical Co., inc. >99.5 %) as the precipitant. After being filtered and washed with hot water, the precipitate was dried at 393 Κ for 12 h, and then pre-calcined in air at 773 Κ for 3 h. Furthermore, they were pressed into disks at 600 kg/m , and then calcined at 1423 Κ for 20 h. The catalysts were crushed and sieved to particles with 150-250 μτη diameter. Methane reforming was carried out in a fixed and a fluidized bed flow reaction systems undo atmospheric and pressurized conditions. The illustration of the fluidized bed reactors is shown in Figure 1. The fluidized bed reactor under atmospheric pressure was the quartz tube (15 mrn^i.d.) with a sintered quartz mesh as a distributor (Figure 1(a)). In the fixed bed reactor, quartz wool was put on the catalyst bed so as to prevent catalysts from moving. Pretreatrnent of catalysts was H reduction at 1173 Κ for 0.5 h under atmospheric pressure. Oxygen was introduced to the reactor through the thin quartz tube, whose outlet was located just before the distributor. C 0 and CH4 were introduced outside the oxygen-feed tube. The partial pressure of the reactant gases was described in each result, and the total pressure was 0.1 MPa. Reaction temperature was monitored inside and outside the reactor. The reaction temperature was controlled by monitoring the thermocouple at the inside. The fluidized bed reactor under pressurized conditions was the quartz tube (6 mm^i.d.) which was placed inside the stainless steel tube (10 mm^i.d.) (Figure 1(b)). A sintered quartz mesh was used as a distributor in fluidized bed reactor. In the fixed bed reactor, quartz wool was put on catalyst bed so as to prevent catalysts from moving. Pretreatrnent of catalysts was H reduction at 1173 Κ for 0.5 h at atmospheric pressure. CH4 was introduced to the reactor through the thin quartz tube, whose outlet was located just before the distributor. C 0 and 0 were introduced into the reactor outside the CH4 feed tube. The conversion of oxygen was 100% in all reaction results. GHSV is calculated on the basis of total gas flow rate of the reactants (CHi+C0 +0 ) at room temperature and under atmospheric pressure. The effluent gas was analyzed with FID gaschromatograph (Gaskuropack 54) equipped with a methanator for CH4, CO, C 0 and T C D gaschromatograph (Molecular Sieve 13X) for H . A n ice bath was set between the reactor exit and a sampling port for G C analysis in order to remove water from the effluent gas. CH4 (99.9%), 0 (99 %), C 0 (99.9 %), and H (99 %) were purchased from Takachiho Co. Ltd., and were used without further purification. x

x

3

3

2

2

2

2

2

3

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

2

-

2

2

2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

306

Figure L Illustration of the fluidized bed reactors, (a) atmospheric pressure condition, (b) pressurized condition.

Resutls and Discussion

Reforming under Atmospheric Pressure Condition Figure 2 shows the dependence of conversion and H /CO ratio on the time on stream in the reforming of methane with C 0 and 0 over Nio.03Mgo.97O catalyst using fixed and fluidized bed reactors under atmospheric reaction condition. In fluidized bed reactor, high CH4 and C 0 conversions were maintained. On the other hand, the conversion decreased gradually with time on steam in the fixed bed reactor. The state of the catalyst in the reactors was observed after the reaction. In the fixed bed reactor, the catalyst near the inlet of the bed was green and the catalyst at the upper part was gray. The green catalyst is in the oxidized state and this indicates that oxygen can reach the green region. In contrast, the gray catalyst is in the reduced state. It has been reported that Nio.03Mgo.97O in oxidized state exhibited no reforming activity. 2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

307

100

1.5

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

75

I

50

ο Ο Ο *3

25

Ο Ο

0.5

U

-25 10

20

30

40

50

60

70

Time on stream/min Figure 2. Comparison of CE4 ( β , Π,) and CO2 (Φ,Ο) conversion andU2/CO (A, A) betweenfluidized(Μ,Φ,Α) andfixedbed {Ο,Ο,Δ^ reactors over Nio.03Mgo.97O. Reaction conditions: reaction temperature 1123 K, total pressure OA MP a, CH/C0 /0 =35/35/30, GHSV=19000 cm/gh, 0.5 g-cat. 2

2

1.5

Η

1 Ο ρ

0.5

1050

1100

1150

1200

Temperature/K Figure 3. Effect of reaction temperature on CE4 (H), CO2 (9) conversion and H2/CO ratio (A) over Nio.03Mgo.97O. Reaction conditions: total pressure 0.1 MPa, CH/CO /O =40/40/20, GHSV=19000 cm /gh, 05 g-cat, fluidized bed reactor. Dotted line represents methane conversion due to the combustion. 3

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

308 It is suggested that the methane conversion decreased with the time on stream because the oxidized region became larger and larger. After the reaction the catalyst in the fluidized bed reactor was almost gray. This indicates that Nio.03Mgo.97O which is oxidized at the inlet of the catalyst bed can be reduced rapidly by produced syngas at the upper part of the catalyst bed. Figure 3 shows the reaction temperature dependence of the conversion and H / C O ratio over Nio.03Mgo.97O using fluidized bed reactor. The reforming reaction proceeded at 1123 and 1173 K , where methane conversion reached about 100%. However, combustion was the main reaction at 1073 K . In this case, the rapid deactivation of the catalyst due to the catalyst oxidation was observed in the fluidized bed reactor. This is contrastive to the fact that the deactivation in the fixed bed is not so rapid as shown in Figure 2. Under the reaction conditions where the rate of the catalyst oxidation is faster than that of the catalyst reduction, the catalyst fluidization drastically enhances the deactivation rate.

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

2

Figure 4. Dependence ofCH (M), C0 (Φ) conversion, H /CO ratio (+), and temperature difference (A) on space velocity over Nio.03Mgo.97O. Reaction conditions: reaction temperature 1123 K, total pressure 0.1 MP a, CHVCO /O =35/35/30, 0.5 g-cat, fluidized bed reactor. Temperature difference = Τ (outside)-T(inside), and dotted line represents methane conversion due to the combustion. 4

2

2

2

2

The difference in methane conversion between experimental results and combustion can be assigned to methane reforming and C O production. Furthermore since H / C O ratio is close to one in the results, the reforming 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

309

reaction is similar to dry reforming. At low temperature, methane conversion was close to that due to combustion, and the main product containing hydrogen is water. At high temperature, methane conversion reached 100%, the selectivity of hydrogen formation can be estimated to be 75%. Figure 4 shows the dependence of the conversion, H / C O ratio, and temperature difference on space velocity over Nio.03Mgo.97O. Methane conversion was high in the space velocity range of 9000-19000 cm /gh. Temperature difference between the reactor inside and outside was almost zero at these space velocities. However, at 27000 cm /gh, methane conversion decreased to the combustion level, and the large temperature difference was observed. A t this condition, only methane combustion proceeded and thus the inside temperature was about 30 Κ higher than the outside temperature. This indicates that the catalyst amount in the oxidized state increased and exhibited no reforming activity at high space velocity. This is probably because the rate of oxidation is faster than that of the reduction of the catalyst. 2

3

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

3

0

10

20

30

0 concentration/% 2

Figure 5. Effect of oxygen concentration on CH4 (M), C0 (%) conversion, H2/CO ratio (Φ), and temperature difference (A) over Nio.03Mgo.97O. Reaction conditions: reaction temperature 1123 K, total pressure 0.1 MP a, CH/CO /O =(50-x)/(50-x)/2x (x=5, 10, and 15), GHSV=19000 cm/gh, 0.5 gcat, fluidized bed reactor, temperature difference = Τ(outside)-T(inside). 2

2

2

Figure 5 shows the effect of oxygen concentration in the reactant on conversion, H / C O ratio, and temperature difference. At 19000 cm /gh, the catalyst exhibited high methane conversion on all oxygen concentration. Temperature difference between the reactor inside and outside in methane 3

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

310 reforming with only C 0 was rather large (about 50 K). The addition of oxygen to the reactant gases made the temperature difference smaller. This is the effect of internal heat supply by the methane combustion. For the stable operation of methane reforming with C 0 and 0 over Nio.03Mgo.97O using the fluidized bed reactor, the space velocity of the reactant gases and the reaction temperature are key factors. This is the relation to the oxidizing and reducing atmosphere in the reactor. 2

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

2

2

Reforming under Pressurized Condition

Figure 6. Dependence ofCH4 (B), CO2 (Φ) conversion andH2/CO ratio (A) on the reaction pressure over Nio.03Mgo.97O. Reaction conditions: reaction temperature 1123 K, total pressure 0.1 MPa, CH4/CO /O =50/25/25 GHSV=37000 cm /gh, 0.3 g-cat,fixedbed reactor. The reactant gas was introduced without using thin tube. 3

2

2

}

Figure 6 shows the dependence of conversion and H /CO ratio on the total pressure over Nio.03Mgo.97O. At atmospheric pressure, methane combustion only proceeded because of the high space velocity. In contrast, under pressurized condition, methane reforming proceeded. Methane conversion decreased with the increase of the total pressure since the equilibrium conversion of methane is lower under higher reaction pressure. In each experiment, the methane conversion was very stable as a function of time on stream. 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

311

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

00

ι

ι

Q

ι

0>

c

,

(b)>(a). In contrast, the difference in methane conversion is so small in the reaction under atmospheric pressure. This suggested that homogeneous methane oxidation proceeds in the gas phase before the reactants reach the catalyst bed. This reaction proceeds more rapidly at higher pressure. Therefore, the amount of oxygen which reaches the catalyst bed in (a) must be smaller than that in (b) and (c). The stable operation was possible in (b). However, it was not in (c). The reaction temperature was not stable and often suddenly increased like the explosion. This is probably because the oxygen concentration was too high at the outlet of the introduction tube. Therefore, we used the reactor as shown in Figure 1(b).

Downloaded by STANFORD UNIV GREEN LIBR on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch020

2

0

30

60

90

120

3

GHSVcm /gh Figure 8. Dependence of CH4 conversion on space velocity usingfluidizedbed (Μ,φ) andfixed bed (Ώ,Ο) reactor over Ni .o3Mgo.970(U,0) and Nio.10Mgo.90O (Φ,Ο).Reaction conditions: reaction temperature 1073 K, total pressure 10 MP a, CH4/CO2/O2S0/20/30, 02 g-cat 0

Figure 8 shows the dependence of methane conversion on the space velocity over Nio.03Mgo.97O and Ni .ioMg .9oO. Under pressurized condition, the conversion was stable at low reaction temperature (1073 K ) . Methane conversion decreased with the increase in the space velocity when the fixed bed 0

0

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

313 reactor was used since the contact time became shorter. Methane conversion was almost the same at space velocity 2CO+2H„

2

CH,+H 0--*CO+3HL 0

H

CO+H Q

2^°2

2

catalyst particle £ ~ cT f>o :o T o o

0

0

reducing atmosphere: Ni (activeforreforming)

ρο

. o'

' ο ο ο · ο. ; /CH range of thermo-neutral state at 40 atm is 0.06 - 0.18 when C 0 / C H changes from 1 to 5. However, at 1 atm, the 0 / C H range of thermo-neutral state is 0.4 - 0.5 depending the C 0 / C H ratio in feedstock. This difference is obviously related to the effect of high pressure on equilibrium composition. 2

4

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch021

O2/CH4 ratio

4

2

4

4

2

4

2

2

4

4

Summary By computational analysis of the reaction system of C 0 and oxy-C0 reforming of methane, we have investigated the effects of reaction temperature, pressure, and feedstock composition on the energy balance, namely the reaction heat needed as energy input for the reaction or released as energy output. Effects of pressure and temperature on the overall energy balance of the reaction system are closely related to their effects on equilibrium composition. Addition of oxygen in the feedstock could turn the global reaction from an endothermic system to an exothermic one. The 0 concentration in feedstock which could lead to the thermo-neutral state in the reaction system strongly depends on the pressure, and it could be calculated and compared under various conditions although only some examples are illustrated in this paper. These data could be very useful for future design of processing scheme and reactor design. 2

2

2

Acknowledgement We are pleased, to acknowledge the partial financial support of this work by the Pennsylvania State University (PSU) and the partial support on C 0 reforming by the Air Products and Chemicals Inc. (APCI). We are grateful to Dr. John N. Armor of APCI, to Prof. Alan W. Scaroni and Prof. Harold H. Schobert of PSU for their encouragement and helpful discussions, to the members of the APCI/PSU SAMCOM for their support, and to Dr. Srinivas T. Srimat for technical assistance and helpful discussions. 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch021

327

-400

J

0 /CH ratio 2

4

Fig. 5 Effect of C 0 / C H and 0 / C H ratio in feedstock on global reaction heat in the reaction systems of C 0 or oxy-C0 reforming of methane at different pressures (upper group: 1 atm, lower group: 40 atm, both reactants and products are at 702°C) 2

4

2

2

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

328 Literature Cited

Downloaded by UNIV MASSACHUSETTS AMHERST on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch021

1. 2.

Armor J.N., Appl. Catal. A: General 1999, 176, 159. Choudhary V.R.; Rajput A.M.; and Prabhakar B. Angew. Chem. Int. Ed. Engl. 1994, 33, 2104. 3. (a) Song, C. Chemical Innovation (formerly Chemtech, ACS), 2001, 31, 21; (b) Song C. Proc. 16th International Pittsburgh Coal Conference: Pittsburgh, USA, October 11-15, 1999, Paper No. 16-5, CD-ROM ISBN 1890977-16-0. 4. Gunardson, H.H.; Abrardo, J.M. Hydrocarbon Processing April 1999, 87. 5. Song, C.; Srinivas, S.T.; Sun, L.; Armor, J.N. Am. Chem. Soc. Div. Petrol. Chem. Prepr., 2000, 45, 143. 6. Srinivas, S.T.; Song, C. Am. Chem. Soc. Div. Petrol. Chem. Prepr., 2000, 45, 153. 7. Pan, W.; Srinivas, T.S.; Song, C. 16th International Pittsburgh Coal Conference: Pittsburgh, USA, October 11-15, 1999; Paper No. 26-2. CD-ROMISBN 1-890977-16-0. 8. Choudhary, V.R.; Rajput, A.M.; Prabhakar, B., Catal. Lett. 1995, 32, 391. 9. O'Connor, A.M.; Ross, J.R.H. Catal. Today, 1998, 46, 203. 10. Ruckenstein, E. and Hu, Y.H. Ind. Eng. Chem. Res., 1998, 37, 1744. 11. Song, C.; Murata, S.; Srinivas, S.T.; Sun, L.; Scaroni, A.W. Am. Chem. Soc. Div. Petrol. Chem. Prepr., 1999, 44, 160.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 22

Photocatalytic Reduction ofCO with H O on Various Titanium Oxide Catalysts Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

2

2

Hiromi Yamashita , Keita Ikeue, and Masakazu Anpo *

*

Department of Applied Chemistry, Graduate School of Engineering, Osaka Prefecture University, Gakuen-cho 1-1, Sakai, Osaka 599-8531, Japan

The characteristic features of the photocatalytic reduction of CO with H O on various types of active titanium oxide catalysts has been clarified. UV-irradiation of active titanium oxide catalysts in the presence of CO and H O at 275 Κ led to the photocatalytic reduction of CO . The reactions on TiO powders produced CH4 as the major product while, on the highly dispersed titanium oxide anchored on porous glass and zeolites, the formations of CH OH as well as CH were observed as the major products. The CH OH formation is linked to unique properties of the charge transfer excited state, i.e., (Ti —O )* of the tetrahedral coordinated titanium oxides species. 2

2

2

2

2

2

3

4

3

3+

-

Introduction The design of highly efficient and selective photocatalytic systems that work without any loss in the use of solar energy through chemical storage like the natural plant photosynthesis is of vital interest (1-8). Especially, the development of efficient photocatalytic systems which are able to reduce CO2 with H 2 O into chemically valuable compounds such as C H 4 or CH3OH are among the most desirable and challenging goals (2,3). The utilization of solar energy for the reduction and/or fixation of CO2 can

330

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

331

be made possible by considering the photocatalytic reduction and/or fixation of CO2 with H2O into C O , HCOOH, CH3OH, and CH4, etc. using reactive photocatalysts such as small particle powdered T1O2 semiconductors. Inoue and Fujishima, et al (9) have first reported that H C O O H , HCHO, and CH3OH are produced by the reduction of CO2 with H2O under irradiation of aqueous suspension systems involving a variety of semiconductor powders such as T1O2 and SrTi03. Although the pioneering works on the photoreduction of CO2 on semiconductors in aqueous suspension systems were summarized by Halmann (10), the efficiency of CO2 reduction was low when water was used as the reductant. Recently, we have reported that these photocatalytic reactions successfully proceed in solid-gas systems on powdered Τ1Ό2 and highly dispersed titanium oxide catalysts at room temperature (2,3,11-22). With these objectives of research in mind, in this chapter, we will focus on the photocatalytic reduction and/or fixation of CO2 with H2O on various types of active titanium oxide catalysts using heterogeneous gas-solid photocatalytic reaction systems. Studies have been carried out on extremely small T1O2 particles and on highly dispersed anchored titanium oxide catalysts under U V irradiation (11-22). C H , C H O H , H C H O , CO 4

3

H 0 2

Fig. 1. Reaction schem efor the photocatalytic reduction of CO2 with H2O on bulk T1O2.

Photocatalytic Reaction on Various Titanium Oxide Catalysts The efficiency for the photosynthetic reduction of CO2 with H2O to produce CH4 and CH3OH depends strongly upon the types of the photocatalyst used (11-22). Among photocatalysjs, the semiconducting T1O2 is the most reactive and stable. Figure 1 shows the primary processes for the photocatalytic reduction and/or fixation of CO2 with H2O on semiconducting T1O2 photocatalysts. When the particle size of the semiconducting T1O2 is decreased, the band gap between the conduction band and the valence band becomes larger, making it suitable and applicable for the reduction of CO2 (2,3). For extremely

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

332 fine particle T1O2 photocatalysts less than 100 Â of the particle, the size quantum effect and/or the effects of the surface modification in its coordination geometry plays a significant role in the reactivity of the photocatalyst (2,3). As a result, the electrons and holes which are produced by UV-irradiation within the ultrafine particles of T1O2 and the highly dispersed titanium oxide species exhibit more unique and high reactivities than for those produced in large particle T1O2 photocatalysts.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

Small particle T1O2 catalysts

UV-irradiation of the powdered T1O2 catalysts in the presence of a gaseous mixture of CO2 and H2O led to the evolution of CH4 into the gas phase at 275 Κ (13,14). Trace amounts of C2H4 and C2H6 were also produced. The yields of these products increased with the UV-irradiation time, while no products were detected under dark conditions. The CH4 yield was almost zero in the reaction of CO2 without H2O and increased when the amount of H2O was increased. These results suggest that for powdered T1O2 catalysts, the photocatalytic reduction of CO2 to produce CH4 and C2-compoundsfromCO2 and H2O takes place photocatalytically in the solid-gas phase systems. The formation of CH4 as the main product has also been recently observed by Saladin et al., and the partially reduced Τΐθ2-δ species formed under UVirradiation is considered to be the active species (23). The yields of CH4 formation in the photocatalytic reduction of CO2 with H2O on several different types of T1O2 catalyst are shown in Table 1. The level of photocatalytic reactivity, based on the CH4 yields, was found to depend on the type of T1O2 catalyst, in the order of JRC-TIO-4 > -5 > -2 > -3 ( 14). As shown in Table 1, the tendency for catalytic activity for the reduction of CO2 with H2O is in a manner similar to those for the hydrogénation of olefins (24,25), proving that the reduction of CO2 with H2O occurs photocatalytically over the powdered T1O2 catalyst. It is likely that the anatase-type T1O2 which has a large band gap and numerous surface -OH groups is preferable for efficient photocatalytic reactions. The band gap increase is accompanied by a shift in the conduction band edge to higher energy levels. This shift causes the reductive potential to shift to more negative values which in turn causes a great enhancement in the photocatalytic reactivity. Concerning the role of the surface -OH groups, the surface -OH groups and/or physisorbed H2O play a significant role in photocatalytic reactions via the formation of OH radicals and H radicals. Figure 2 shows the ESR signals obtained under UV-irradiation of the anatase-type T1O2 catalyst in the presence of CO2 and H2O at 77 Κ (14). The ESR signals are attributed to the characteristic photogenerated T p ions (gj_= 1.9723 and g//= 1.9628) and Η radicals (with 490 G splitting), as well as CH3 radicals having a hyperfinesplitting (Ha=19.2 G, g=2.002). The signal intensity of CH3 radicals decreased with increasing the amount of H2O, indicating that CH3 radicals react easily to form CH4 in the presence of H2O. +

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

333 These results clearly suggest that CH3 radicals are the intermediate species and react with H radicals that are formed by the reduction of protons (H ) supplied from H2O adsorbed on the catalyst. +

Table 1. Physical properties and photocatalytic activity of T1O2 catalysts. (CO2: 0.12 mmol, H2O: 0.37 mmol, irradiation time:6 hr, ). Surface Catalyst area (JRC-TIO-) (m g' )

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

2

2 (anat.)

1

Relative -OH cone.

16

Band gap (eV)

1

3.47

3

Reduction* ofC0 (μιηοΐ h' g' ) 2

1

1

Hydrogénation^ of olefins (μιηοΐ h" g' ) 1

0.03

0.20

3 (ruti.)

51

1.6

3.32

0.02

0.12

4 (anat)

49

3.0

3.50

0.17

8.33

5 (ruti)

3

3.1

3.09

0.04

0.45

1

" CH4 yield in reaction of CO2 with H2O. Conversion in reaction of methyl acetylene with H2O. h

CH

3

radicals

H radicals

1 f

100 G g//=1.962

8

Fig. 2. ESR signals obtained with T1O2 under UV-irradiation in the presence of CO2 and H2O at 77 K.

Cu-loaded and Pt-loaded T1O2 powder The effect of Cu-loading on the photocatalytic reduction of CO2 with H2O on T1O2 catalysts was investigated (14). Although the loading of Cu (0.3-1.0 wt %) onto the small particle T1O2, i.e., the CU/T1O2 photocatalyst, led to a suppression of the CH4 yield, a new formation of CH3OH could be observed. Characteristics of the XPS spectra observed with the CU/T1O2 suggest that the main species of copper in the catalyst is C u . It has been also reported that C u catalysts play a significant role in the photoelectrochemical production of CH3OH from CO2 and H2O system (26). +

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

+

334 The effect of Pt-loading on the T1O2 catalyst was also investigated. The yield of CH4 increased remarkably when the amount of Pt added was increased (0.1-1.0 wt%), but the addition of excess Pt was undesirable for an efficient reaction (13). Regarding the reaction intermediates, Solymosi et al. have observed the formation of C O 2 ' species in bent form under UV-irradiation of the Rh/Ti02 the presence of CO2 using FT-IR (27). The electron transfer from the irradiated catalyst to the adsorbed CO2 takes place, resulting in the formation of a CO2" anion as the key step in the photochemical reduction of CO2 using metal loaded T1O2 semiconductor. Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

m

T1O2 single crystals With a well-defined catalyst surface such as a single crystal, detailed information on the reaction mechanism can be obtained at the molecular level (2,15). Therefore, the photocatalytic reduction of CO2 with H2O on rutile-type TiO2(100) and TiU2(l 10) single crystal surfaces have been carried out (15). As shown in Table 2, the efficiency and selectivity of the photocatalytic reactions strongly depend on the type of T1O2 single crystal surface. UV-irradiation of the TiO2(100) single crystal catalyst in the presence of a mixture of CO2 and H2O led to the evolution of CH4 and CH3OH in the gas phase at 275 K, whereas only CH3OH yields was detected with the TiO2(H0) single crystal catalyst.

Table 2. Yields of the formation of CH4 and CH30H in the photocatalytic reduction of CO2 (124 μπιοί g") with H2O (372 μπιοί g") at 275 Κ. 1

Single Crystal

1

1

1

Yields (nmol h " g-cat" ) CH

4

CH3OH

TiO (100)

3.5

2.4

T i 0 (110)

0

0.8

2

2

It is likely that the photo-formed electrons localize on the surface sites of the excited T1O2 to play a significant role in the photoreduction of CO2 molecules into intermediate carbon species (2,15). The surface Ti atoms may act as an electron moiety on the surfaces, i. e., a reductive site. According to the surface geometric models for TiO2(100) and TiU2(l 10), the atomic ratio (Ti/O) of the top-surface Ti and Ο atoms which have geometric spaces large enough to have direct contact with CO2 and H2O molecules, is higher on TiO2(100) than on Ti02(l 10) surface. In the excited state, the surface with a higher Ti/O surface ratio, i. e. TiO2(100), exhibits a more reductive tendency than Ti02(l 10). Such

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

335 a reductive surface allows a more facile reduction of CO2 molecules especially for the formation of CH4.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

Ti-oxide anchored on zeolite (ion-exchange) The photocatalyst systems incorporated within the zeolite cavities and framework have been found to be effective for various reactions (28-31). The Ti-oxide anchored onto zeolite, Ti-oxide/Y-zeolite (1.1 wt% as T1O2), was prepared by ion-exchange with an aqueous titanium ammonium oxalate solution using Y-zeolite (S1O2/AI2O3 = 5.5) (ex-Ti-oxide/Y-zeolite) (16-18).

Λ Ti-O-Ti / 1

s~\

/ 1/

\

Ν = 6.0

\

(b) R = 1.93 Ν = 5.8

A

/

/

\

A

^

(a) R = 1.93

; A Ti-0 -

-

-/ 4920

4960

5000

Energy (eV )

5040

0

/ \ 1 \

(c) R = 1.78 Ν = 3.7

2

4

6

Distance (Â)

Fig. 3. TiK-edgeXANES (A-C) and FT-EXAFS (a-c) spectra of anatase T1O2 powder (A, a), the imp- Ti-oxide/Ύ-zeolite (10.0 wt% as T1O2) (B, b), and the ex-Ti-oxide/Y-zeolite (C, c),catalysts. R: Ti-O bond distance (À), N: coordination number Figure 3 shows the X A N E S and the Fourier transformation of E X A F S (FT-EXAFS) spectra of the Ti-oxide/Y-zeolites. The X A N E S spectra of the T i oxide catalysts at the Ti K-edge show several well-defined preedge peaks which are related to the local structures surrounding the Ti atom (32). The ex-Tioxide/Y-zeolite exhibits an intense single preedge peak indicating that the T i oxide species have a tetrahedral coordination (1-3). On the other hand, the impTi-oxide/Y-zeolite prepared by the impregnation exhibits three characteristic weak preedge peaks attributed to crystalline anatase T1O2. The F T - E X A F S spectra of the ex-Ti-oxide/Y-zeolite exhibits only peak assigned to the neighboring oxygen atoms (Ti-O) indicating the presence of an isolated T i oxide species. These findings indicate that highly dispersed isolated tetrahedral Ti-oxide species are formed on the ex-Ti-oxide/Y-zeolite. On the other hand, the

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

336 imp-Ti-oxide/Y-zeolite exhibits an intense peak assigned to the neighboring titanium atoms (Ti-O-Ti), indicating the aggregation of the Ti-oxide species. Figure 4a shows a typical photoluminescence spectrum of the Ti-oxide anchored onto zeolite (ex-Ti-oxide/Y-zeolite) at 77 K. Excitation by light at around 250-280 nm brought about an electron transfer from the oxygen to titanium ion, resulting in the^ formation of pairs of the trapped hole center (O") and an electron center ( T i ^ ) (1-3). The observed photoluminescence is attributed to the radiative decay process from the charge transfer excited state of the Ti-oxide moieties having a tetrahedral coordination, (Ti^ —O")*, to their ground state (33,34). A s shown in Fig. 4b,c, the addition of H2O or CO2 molecules onto the anchored Ti-oxide species leads to the efficient quenching of the photoluminescence. Such an efficient quenching suggests not only that tetrahedrally coordinated Ti-oxide species locate at positions accessible to the added CO2 or H2O but also that added CO2 or H2O interacts and/or reacts with the Ti-oxide species in both its ground and excited states. Because the addition of CO2 led to a less effective quenching than with the addition ofH2Û, the interaction of the emitting sites with CO2 was weaker than with H2O. +

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

+

200

300

400 500 Wavelength (nm)

600

Fig. 4. Photoluminescence spectrum of the ex-Ti-oxide/Y-zeolite catalyst (a), its excitation spectrum (EX), and the effects of the addition of CO2 and H2O (b, c) and the loading of Pt (d) on the photoluminescence spectrum. Measured at 77 K, excitation at 290 nm, emission monitored at 490 nm, amounts of added CO 2- b) 8.5, and H2O; c) 2.9 mmol g~K UV-irradiation of powdered T1O2 and Ti-oxide/Y-zeolite catalysts in the presence of a mixture of CO2 and H2O led to the evolution of CH4 and CH3OH in the gas phase at 328 K , as well as trace amounts of CO, C2H4 andC2H6(1618). The yields of these photoformed products increased linearly against the UV-irradiation tnrie, indicating the photocatalytic reduction of CO2 with H2O on the catalysts. The specific photocatalytic reactivities for the formation of CH4 and CH3OH are shown in Fig. 5. The ex-Ti-oxide/Y-zeolite exhibits a

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

337 high reactivity and a high selectivity for the formation of CH3OH while the formation of CH4 was found to be the major reaction on bulk T1O2 as well as on the imp-Ti-oxide/Y-zeolite. These findings clearly suggest that the tetrahedrally coordinated Ti-oxide species act as active photocatalysts for the reduction of CO2 with H2O showing a high selectivity to produce CH3OH.

-

15

!

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

Q σ> 10

•

CH

M

C

H

4

3°

H

ô

ε

>

ο

(a)

Π (b).

(c) Catalysts

(d)

(β)

Fig. 5. The product distribution of the photocatalytic reduction of CO2 with H2O on anatase T1O2 powder (a), the imp-Ti-oxide/Y-zeolite (10.0 wt% as T1O2) (b), the imp-Ti-oxide/Y-zeolite (L0wt%as T1O2) (c), the ex-Ti-oxide/Yzeolite (1.1 wt% as T1O2) (d), and the Pt-loaded ex-Ti-oxide/Y-zeolite (e) catalysts. UV-irradiation of the anchored Ti-oxide catalyst in the presence of CO2 and H2O at 77 Κ led to the appearance of ESR signals due to the T i ^ ions, Η atoms, and carbon radicals (2,3). From these results, the reaction mechanism in the photocatalytic reduction of CO2 with H2O on the highly dispersed Ti-oxide catalyst can be proposed: CO2 and H2O molecules interact with the excited state of the photoinduced ( T p — Ο " ) * species and the reduction of CO2 and the decomposition of H2O proceed competitively. Furthermore, Η atoms and Ο Η · radicals are formed from H2O and these radicals react with the carbon species formed from CO2 to produce CH4 and CH3OH. +

+

Ti-containing zeolite and mesoporous molecular sieves The Ti-oxide species prepared within the zeolite framework have revealed a unique local structure as well as a high selectivity in the oxidation of organic substances with hydrogen peroxide (33-35). Ti-containing zeolites (TS1, Ti-Beta) and mesoporous molecular sieves ( T i - M C M , Ti-HMS, Ti-FSM) have been hydrothermally synthesized (18-22, 35-37).

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

338 In situ photoluminescence, ESR, UV-VIS and XAFS investigations indicated that the Ti-oxide species in the Ti-mesoporous molecular sieves (TiMCM-41 and Ti-MCM-48) and in the TS-1 zeolite are highly dispersed within the zeoliteframeworkand exist in atetrahedral coordination. Upon excitation with U V light at around 250-280 nm, these catalysts exhibit photoluminescence spectra at around 480 nm. The addition of CO2 or H2O onto these catalysts results in a significant quenching of the photoluminescence, suggesting the excellent accessibility of the Ti-oxide species to CO2 and H2O (18-21). UV-irradiation of the Ti-mesoporous molecular sieves and the TS-1 zeolite in the presence of CO2 and H2O also led to the formation of CH3OH and CH4 as the main products (18-21). The yields of CH3OH and CH4 per unit weight of the Ti-based catalysts are shown in Fig. 6. It can be seen that TiMCM-48 exhibits much higher reactivity than either TS-1 or Ti-MCM-41. Besides the higher dispersion state of the Ti-oxide species, other distinguishing features of these zeolite catalysts are: TS-1 has a smaller pore size (ca. 5.7 Â) and a three-dimensional channel structure; Ti-MCM-41 has a large pore size (>20 Â) but one-dimensional channel structure; and Ti-MCM-48 has both a large pore size (>20 Â) and three-dimensional channels. Thus, the higher reactivity and higher selectivity for the formation ofCH30H observed with the Ti-MCM-48 than with the other catalysts may be due to the combined contribution of the high dispersion state of the Ti-oxide species and the large pore size with a three-dimensional channel structure. These results strongly indicate that mesoporous molecular sieves with highly dispersed Ti-oxide species are promising candidates as effective photocatalysts.

(b)

(c)

(d)

Catalysts

Fig. 6. The product distribution of the photocatalytic reduction of CO2 with H2O on Ti0 powder (a), TS-1 (b), Ti-MCM-41 (c), Ti-MCM-48 (d), and the Pt-loaded Ti-MCM-48 (e) catalysts. 2

The effect of Pt-loading on the photocatalytic reactivity of Ti-containing zeolite has also been investigated and the changes in the yields of CH4 and

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

339

CH3OH formation are shown in Figs. 5 and 6. Although the addition of Pt onto the Ti-containing zeolites is effective for an increase in the photocatalytic reactivity, only the formation of CH4 is promoted (16,18). Recently a large-pore Ti-containing zeolite, Ti-Beta, has been hydrothermally synthesized (35-37). The H2O affinity of Ti-Beta zeolites changes significantly depending on the preparation methods and their hydrophobic-hydrophilic properties can modify the catalytic properties (35-37). As shown in Fig. 7, the photocatalytic reduction of CO2 with H2O to produce CH4 and CH3OH was found to proceed in the gas phase at 323 Κ with different reactivities and selectivities on hydrophilic Ti-Beta(OH) and hydrophobic T i Beta(F) zeolites prepared in the OH" and F" media, respectively. The higher reactivity for the formation of CH4 observed with Ti-Beta(OH) and the higher selectivity for the formation of CH3OH observed with the Ti-Beta(F) may be attributed to the different abilities of zeolite pores on the H2O affinity. These results suggest that the affinity of the H2O molecules to adsorb on the zeolite is one of important factors for the selectivity in the photocatalytic reduction of CO2 and H2O.

P-25

Ti-Beta(F)

Ti-Beta(OH)

Catalysts

Fig. 7. The product distribution of the photocatalytic reduction of CO2 with H2O on Ti-Beta(F), Ti-Beta(OH), and T1O2 powder (P-25) as the reference catalyst. In summary, an efficient photocatalytic reactivity and selectivity for the formation of CH3OH in the photocatalytic reduction of CO2 with H2O was achieved with the zeolite and mesoporous molecular sieves having highly dispersed tetrahedral Ti-oxide species in their cavities or frameworks, while the formation of CH4 in the photocatalytic reduction of CO2 with H2O was found to proceed on the bulk T1O2 catalysts and on the catalysts involving aggregated Ti-oxide species.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

340

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

Ti-oxide anchored on porous silica glass (CVD) The Ti-oxide anchored onto porous silica glass (PVG) plate was prepared using a facile reaction of T1CI4 with the surface OH groups on the transparent porous Vycor glass (Corning code 7930) in the gas phase at 453-473 K, followed by treatment with H2O vapor to hydrolyze the anchored compound (11-13). UV-irradiation of the anchored Ti-oxide catalysts in the presence of a mixture of CO2 and H2O led to the evolution of CH4, CH3OH and CO in the gas phase at 323 K. The total CH4, CH3OH and CO yields were larger under UV-irradiation at 323 Κ than at 275 K. The efficiency of the photocatalytic reaction strongly depends on the ratio of H2O/CO2 and its reactivity increases with an increase in theH20/C02 ratio; however, an excess amount of H2O suppresses the reaction rates. Figure 8 shows the effect of the number of anchored Ti-O layers on the absorption edge of the catalysts and the efficiency of the photocatalytic reactions as well as the relative yields of the photoluminescence (11-13). It was found that only catalysts with highly dispersed monolayer Ti-oxide exhibit high photocatalytic reactivity and photoluminescence at around 480 nm. Only the tetrahedral Ti-oxide species exhibit photoluminescence when it is excited at around 250-280 nm. These findings also clearly suggest that the tetrahedrally coordinated Ti-oxide species act as active photocatalysts for the reduction of CO2 with H2O.

Number of Ti-0 layer Fig. 8. The effects of the number of the Ti-O layers of the anchored titanium oxide catalysts on the absorption edge of the catalysts (a), the reaction yields (b), and the relative yields of the photoluminescence (c).

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

341

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

Ti/Si binary oxide (sol-gel) Ti/Si bmary oxides involving different Ti contents were prepared by the solgel method using mixtures of tetraethylorthosilicate and titanium-iso-propoxide. Ti/Si binary oxides with only a small Ti contents exhibited the photoluminescence at around 480 nm upon excitation at around 280 nm (22, 38). UV-irradiation of the Ti/Si binary oxide catalysts in the presence of a gaseous mixture of CO2 and H2O was found to lead to the formation of CH4 and CH3OH as the main product. As shown in Fig. 9, a parallel relationship between the specific photocatalytic activities of the titanium oxide species and the photoluminescence yields of the Ti/Si binary oxides clearly indicates that the appearance of high photocatalytic activity for the binary oxides is closely associated with the formation and reactivity of the charge transfer excited complex of the highly dispersed tetrahedral Ti-oxide species.

0

20

40

60

80

100

Ti0 content ( mol %) 2

Fig. 9. Effects of TO2 contents on the yields of CH4 and the relative intensity of photoluminescence of the Ti/Si binary oxide catalysts at 77 K. (CO2: 0.07 mmolg' , H2O: 0.33 mmolg' 325 K) 1

1

y

The X A F S , ESR and photoluminescence investigations of the Ti/Si binary oxide indicated that these catalysts prepared by the sol-gel method are able to sustain a tetrahedral coordination of Ti-oxide species until the Ti content reaches approximately up to 20 wt% as T1O2. Thus, we can see that the Ti/Si binary oxides having a high Ti content (up to 20 wt%) can be successfully utilized as active photocatalysts for the efficient reduction of CO2 with H2O in the gas-solid system.

Conclusions The characteristic features of the photocatalytic reduction of CO2 with H2O on various types of active Ti-oxide catalysts has been clarified. UV-irradiation

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

342 of active Ti-oxide catalysts in the presence o f C 0 2 and H2O at 275 Κ led to the photocatalytic reduction of CO2. The reactions on T1O2 powders produced CH4 as the major product, while on the highly dispersed titanium oxide anchored on porous glass and zeolites, the formations of CH3OH as well as CH4 were observed as the major products. The yields of the photocatalytic reactions strongly depended on the type of catalysts, the value of CO2/H2O, and the reaction temperature. In situ spectroscope studies of the system indicated that the photocatalytic reduction of CO2 with H2O is linked to a much higher reactivity of the charge transfer excited state, i.e., (Ti^ —O")* of the tetrahedral coordinated Ti-oxide species formed on the surface. Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

+

References 1. 2. 3.

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

15. 16.

Anpo. M; Yamashita, H. In Surface Photochemistry; Anpo. M. Ed.; John Wiley & Sons; Chichester, 1996, pp. 117-164. Anpo. M; Yamashita, H. In Heterogeneous Photocatalysts; Schiavello. M. Ed; John Wiley & Sons; Chichester, 1997, pp. 133-168. Yamashita, H . ; Zhang, J . ; Matsuoka, M . ; Anpo, M . In Photofunctional Zeolites; Anpo, M . Ed.; N O V A ; New York; 2000, pp. 129-168. Anpo, M.; Yamashita, H.; Zhang, S. G. Current Opinion in Solid State & Materials Science 1996, 1, 219-224. Anpo, M. Res. Chem. Intermed. 1989, 9, 67-112. Anpo, M.; Ichihashi, Y.; Takeuchi, M.; Yamashita, H. Res. Chem. Intermed. 1998, 24, 143-149. Yamashita, H.; Ichihashi, Y.; Takeuchi, M.; Kishiguchi, S.; Anpo, M. J. Synchrotron Rad. 1999, 6, 451-452. Yamashita, H.; Honda, M.; Harada, M . ; Ichihashi, Y.; Anpo, M. J. Phys. Chem. Β 1998, 102, 10707-10711. Inoue, T.; Fujishima, Α.; Konishi S.; Honda, K. Nature 1979, 277, 637-640. Halmann, M. In Energy Resources through Photochemistry and Catalysis; Grätzel M. Ed.; Academic Press, New York, 1983, pp. 507-565. Anpo, M.; Chiba, K. J. Mol. Catal. 1992, 74, 207-302. Yamashita, H.; Shiga, Α.; Kawasaki, S.; Ichihashi, Y.; Ehara, S.; Anpo, M. Energy Conv. Manag. 1995, 36, 617-620. Anpo, M.; Yamashita H.; Ichihashi Y.; Ehara S. J. Electroanal. Chem. 1995, 396, 21-26. Yamashita, H.; Nishiguchi, H.; Kamada, N.; Anpo, M.; Teraoka, Y.; Hatano, H.; Ehara, S.; Kikui, K.; Palmisano, L.; Sclafani, Α.; Schiavello M.; Fox, M. A. Res. Chem. Intermed. 1994, 20, 815-823. Yamashita, H.; Kamada, N.; He, H.; Tanaka, K.; Ehara, S.; Anpo, M. Chem. Lett. 1994, 855-858. Anpo, M . ; Yamashita, H.; Ichihashi, Y.; Fujii, Y.; Honda, M. J. Phys. Chem.B 1997, 101, 2632-2636.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

343 17.

18. 19.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch022

20. 21. 22.

23. 24. 25. 26. 27. 28.

29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

Anpo, M.; Yamashita, H.; Fujii, Y.; Ichihashi, Y.; Zhang, S. G.; Park, D. R.; Ehara, S.; Park, S. E.; Chang, J. S.; Yoo. J. W. Stud. Surf. Sci. Catal. 1998, 114, 177-182. Yamashita, H.; Fuji, Y.; Ichihashi, Y.; Zhang, S. G.; Ikeue, K.; Park, D. R.; Koyano, K.; Tatsumi, T.; Anpo, M. Catal.Today 1998, 45, 221-227. Anpo, M.; Zhang, S. G.; Fujii, Y.; Ichihashi, Y.; Yamashita, H.; Koyano, K.; Tatsumi. T. Catal. Today 1998, 44, 327-332. Ikeue, K.; Yamashita, H.; Anpo, M. Chem. Lett. 1999, 1135-1136. Yamashita, H.; Ikeue, K.; Takewaki, T. Topics in Catal. in press. Yamashita, H.; Kawasaki, S.; Fujii, Y.; Ichihashi, Y.; Ehara, S.; Park, S. E.; Chang, J. S.; Yoo. J. W.; Anpo, M. Stud. Surf. Sci. Catal. 1998, 114, 561564. Saladin, F.; Forss, L.; Kamber, I. J. Chem. Soc., Chem. Commun. 1995, 533534. Anpo, M.; Tomonari, M.; Fox, M. A. J. Phys. Chem. 1989, 93, 7300-7303. Yamashita, H.; Ichihashi, Y.; Harada, M.; Stewart, G.; Fox M. Α.; Anpo, M. J. Catal., 1996, 158, 97-101. Frese, K. W. J. Electrochem. Soc. 1991, 138, 3338-3343. Raskö, J.; Solymosi, F. J. Phys. Chem. 1994, 98, 7147-7152. Anpo, M.; Matsuoka, M.; Shioya, Y.; Yamashita, H.; Giamello, E.; Morterra, C.; Che, M.; Patterson, H. H.; Webber, S.; Ouellette, S.; Fox, M. A. J. Phys. Chem., 1994, 98, 5744-5750. Yamashita, H.; Matsuoka, M . ; Tsuji, K.; Shioya, Y.; Anpo M.; Che, M . J. Phys. Chem. 1996, 100, 397-402. Yamashita, H.; Ichihashi, Y; Anpo M.; Hashimoto, M . ; Louis, C.; Che, M. J. Phys. Chem. 1996, 100, 16041-16044. Yamashita, H.; Zhang, S. G.; Ichihashi, Y.; Matsumura, Y.; Souma, S.; Tatsumi, T.; Anpo, M . Appl. Surf. Sci. 1997, 121, 305-309. Farges, F.; Brown G. E. Jr.; Rehr, J. H. Geochim. Cosmochim. Acta 1996, 60, 3023-3060. Anpo, M; Aikawa, N.; Kubokawa, Y.; Che, M.; Louis, C.; Giamello, E. J. Phys. Chem. 1985, 89, 5017-5021. Anpo, M . ; Aikawa, N.; Kubokawa, Y.; Che, M.; Louis, C.; Giamello, E. J. Phys. Chem. 1985, 89, 5689-5694. Takewaki, T.; Hwang, S. J.; Yamashita, H.; Davis, M. E . Micropor. Mesopor. Mater. 1999, 32, 265-273. Camblor, M. Α.; Corma, Α.; Esteve, P.; Martines M . ; Valencia, S. Chem. Commun. 1997, 795-796. Tatsumi, T.; Jappar, N. J. Phys. Chem. B 1998, 102, 7126-7131. Yamashita, H.; Kawasaki, S.; Ichihashi, Y.; Harada, M.; Anpo, M.; Stewart, G.; Fox, M. Α.; Louis, C.; Che, M.J. Phys. Chem. B 1998, 102, 5870-5875.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 23

Electrochemical Reduction of CO with Gas-Diffusion Electrodes Fabricated Using Metal and Polymer-Confined Nets 2

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

K. Ogura, H. Yano, andM.Nakayama Department of Applied Chemistry, Yamaguchi University, Tokiwadai 2-16-1, Ube 755-8611, Japan

Electrochemical reduction of CO has been performed with a Cu/gas-diffusion electrode, a Cu plate electrode, and a modified Pt/gas-diflusion electrode in KCl (pH 3) and KHCO (pH 7.2) solutions. Among these electrolysis systems, the most efficient reduction of C O was obtained in a KCl solution of pH 3 with a Cu/gas-diffusion electrode on which CO, C H and CH were mainly formed with some quantities of solution products such as lactic acid and ethanol. The hydrogen evolution was sufficiently retarded even at highly negative potential, and the total current efficiency was almost 100 %. On the other hand, the electrochemical reduction of CO was carried out at a Cu plate electrode in a K H C O solution of pH 7.2. This solution has been exclusively used as an electrolyte in the reduction of CO at a Cu plate electrode. Major products were CH and CH, and the total current efficiency was about 43 %. The cause for such a low current efficiency was attributed to the formation of graphitic carbon and/or 2

3

2

2

4

4

2

3

2

4

2

344

4

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

345 copper oxide. At a Prussian blue (PB)/polyaniline (PAn)/metal complex-confined Pt/gas-diflusion electrode in a KCl solution of pH 3, lactic acid was mainly produced, and alcohols and acetic acid were minor products. In this electrolysis system, C O is bifunctionally captured and activated by ΡAn and a metal complex: the electrophilic caibon atom of CO2 is bound to the amino group of ΡAn, and the basic oxygen atom coordinates to the central metal of the complex.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

2

Chemical fixation of carbon dioxide is of importance in connection with the mitigation of the concentration of green-house gas in the atmosphere. Many processes have been proposed for the chemical conversion of C 0 including the hydrogénation over heterogeneous and homogeneous catalysts at high temperature and electrochemical reduction (J). One of the most essential matters in C 0 fixation is to achieve it under an input energy as low as possible to avoid a secondary generation of C 0 . Because of this, an electrochemical reduction process taking place at room temperature seems to be promising. In the application of this method, however, there are still difficult problems to be settled, e. g., the deactivation of electrode in a prolonged electrolysis. There are many works on the electrochemical reduction of C 0 with Cu electrode in a hydrogencarbonate solution (2-9). This electrolysis system is known to produce mainly methane and ethylene. In a prolonged electrolysis of C 0 , however, the rate of formation of hydrocarbons is found to go down to very small value. It is pointed that the high faradaic efficiencies of CH4 and C H4 reported in most of literatures are not the steady-state values but rather maximum efficiencies observed in a short period of electrolysis (8). This is attributed to the formation of a poisoning species. Therefore, it is very difficult to use the electrochemical process with Cu electrode for the reduction of C 0 over long periods. The poisoning of this reaction has been attempted to be alleviated by applying a periodic anodic polarization, and the reactivation of the electrode been confirmed (8). Although a neutral hydrogencarbonate solution is the unique electrolyte in which hydrocarbons are produced on a Cu plate electrode in the reduction of C 0 , this solution becomes alkaline with the progress of the reduction, resulting in less efficiency of C 0 conversion. On the other hand, a Cu plate electrode does not lead to the reduction of C 0 in acidic solution but only to the hydrogen evolution. 2

2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

346 In the present study, we have developed a gas-diffusion electrode consisting of a metal mesh electrode and a glass filter. As described later, the electrolysis of C 0 at the gas-diffusion electrode with Cu mesh electrode led chiefly to the formation of CO and C2H4 with high yields in acidic solution, and the poisoning problem of the electrode was not involved. In the gas-diffusion electrode, C 0 can be always supplied in large quantities to the electrode surface and the hydrogen evolution was effectively suppressed even in acidic solution. In addition to the experiments with Cu plate and mesh electrodes, a platinum mesh substrate on which an inorganic conductor and a conducting polymer were immobilized was also used as an electrode for the C 0 reduction. 2

2

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

2

Experimental The electrolysis cell with a gas-diffusion electrode used (Figure 1) was essentially the same as that reported elsewhere (JO). The working electrodes were a copper mesh, a copper plate and a platinum mesh, and their effective surface areas coming into contact with electrolyte were 10.2 cm , 10.2 cm and 19.3 cm , respectively. Copper mesh and copper plate were made from pure copper (99.9 % purity, Nilaco Co.). The platinum mesh was modified with Prussian blue (PB, KFe^fFe^CNQe]) (inner layer), conducting polymer (PAn) (outer layer), and a metal complex, following the procedure described previously (11). Anions incorporated in the prepared Pt mesh/PB/PAn electrode were released by immersing it in a phosphate buffer of pH 7, and bis(l,8-dihydroxynaphthalene-3,6disulfonato) ferrate(II) (Fe L") complex was electrodeposited from its aqueous solution. As shown in Figure 1, these mesh electrodes were put on a glassfilterthrough an O-ring, and both were bound to a Teflon cylinder tightened with a Teflon screw cup. This was connected to the cathode compartment through an O-ring which was separated from the anode compartment by a cation-exchange membrane. The average pore size of glassfilterwas 20 ju m. Purified C 0 gas was forced up through the glass filter, and the liquid meniscuses spread out on the mesh electrode to form a thin layer electrolyte. The electrolysis with a Cu plate electrode was performed using a Η-type cell. The electrolytes used were 0.5 M KC1 solutions of various pH's and 0.5 M KHCO3 solution of pH 7.2. These solutions were prepared from reagent grade KC1, KHCO3, H Q and K O H with doubly distilled water. The counter electrode was a platinum plate, and the reference electrode was a Ag/AgCl/saturated KC1 electrode, The quantitative analyses of the reaction products were done with a gas 2

2

2

n

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

c

Figure 1. Schematic diagram of the electrolysis cell A, gas inlet; B, gas outlet; C, counter electrode; D, electrolyte; E, working electrode; F, reference electrode; G , glassfilter;H, O-ring; I Teflon screw cap; J , cation-exchange membrane.

Β

F

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

348 chromatograph (Shimadzu GC-8A), a steam chromatograph (Ohkura SSC1) and an organic acid analyzer (Shimadzu LC-LoAD).

Results and Discussion

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

Reduction of C0 Electrodes

with Copper/Gas-Diffusion and Copper Plate

2

The electrochemical reduction of CO2 was performed with Cu/gasdiffusion and Cu plate electrodes in a 0.5 M KC1 solution of pH 3 and a 0.5 M KHCO3 of pH 7.2, respectively, and the results are shown versus potential in Figure 2. These solutions were selected because the yields of CO2 reduction were maximum in each electrolysis system. As seen from this figure, the reduction of C 0 occurs at more negative potential than -1.2 V vs Ag/AgCl on the Cu plate electrode, but on the Cu/gas-diffiision electrode the CO2 reduction is observed even around -0.6 V although the full-scale reduction begins from -1.2 V. The total concentration of the products obtained with the gas-difiusion electrodes was about 3.5 times as large as that with the plate electrode at -2.4 V. Hence, the gas-diffusion electrode is found to be much more effective for the C 0 reduction than the plate electrode. Faradaic efficiencies for the products obtained in the electrochemical reduction of C 0 with a Cu plate electrode and a Cu/gas-diflusion electrode are shown in Table I and Π, respectively. The major products except hydrogen are CH4 and C2H4 on the Cu plate electrode, and the total current efficiency averages about 43 % (Table I). This value is very low compared to that obtained with the Cu/gas-diffiision electrode as looted below. The cause for such a low current efficiency may be attributed to the formation of poisoning species on the Cu plate electrode. The identification of this species is under investigation, but graphitic carbon is one of candidates (5). On the other hand, the major products obtained with the Cu/gas-diffiision electrode are CO, C2H4 and CH4 as shown in Table Π. The total current efficiency was almost 100 %, which is a contrast to the value of about 43 % on the Cu plate electrode. It is therefore indicated that the electrolysis system with the Cu/gas-diffiision electrode is not subject to poisoning. This is probably because the poisoning process is considerably suppressed in acidic solution. Major compounds generated in the two electrolysis systems are shown versus potential in Figures 3 and 4. The formation of CH4 and C2H4 was observed on the Cu plate electrode at more negative potential than -1.4 V, 2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

349

-2.6

-2.2

-1.8

-1.4

-1

-0.6

-0.2

Potential/VvsAg/AgCI Figure 2. Relationship between the total concentration ( Σ of the products based on the C content and the electrode potential on Cu plate (O) and Cu/gas-diffusion ( φ) electrodes in 0.5 M KHCO3 and 0.5 MKCl solutions, respectively. Σ C = [CH OH] + [HCOOH] + [CO] + [CH ] + 2([CHsCHO] + [C H OH] + [CH3COOH] + [C2H4} + [C H

C H4 + 40H"

(6)

2

2

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

2

2

2

The standard potentials for reaction (4) and (5) are known to be 0.206 V vs SHE and 0.132 V vs SHE, respectively. As another possible scheme for the formation of graphite instead of reaction (4), the involvement of copper oxide is considered as following. C 0 + Cu + H 0 + 2e" 2

2

-> C + CuO + 20H"

(7)

In fact, copper oxide has been detected on the Cu plate electrode used for the C 0 reduction in hydrogencarbonate solution (9). The formation of carton on the catalyst surface has been often referral in the reduction of CO over a heterogeneous catalyst at high temperature, and hydrocarbons are produced by hydrogénation of surface carbon (12). The formation of surface carbon has been identified also in the methanation of C 0 (13). However, a striking difference between electrochemical and high temperature reduction of C 0 is that the former reduction on Cu plate electrode is not accompanied by the generation of CO. The cause for such a difference should to clarified by further investigation. 2

2

2

Reduction of C 0 with a Modified Pt/Gas-Diffusion Electrode 2

The electrochemical reduction of C 0 was performed with a modified Pt/gas-diffiision electrode. The modification of a Pt substrate was achieved with an inorganic conductor and a conducting polymer, and besides a metal complex was immobilized to the conducting polymer (10). The results obtained are shown in Table III, where the flow rate of C 0 was changedfrom2.0 to 99.0 ml min . The major product was lactic acid, and methanol, ethanol, acetone, 2-propanol, and formic and acetic acids were formed as minor products. The current efficiency for the reduction of C 0 2

2

1

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002. 2

2.5

5.0 12.7

26.7

20.6

15.1

63.7

57.0

32.0

142

135

134

137

141

1.9

2.2

2.2

3.3

2.3

25.6

44.3

36.6

32.8

16.9

e

2

5

2

Faradaic efficiency for the reduction of C0 .

t

3

X C = [CHjOH] + [HCOOH] + 2([C H OH] + [ C ^ C O O H ] ) + S t f C ^ C O C H J + [CH Œ(OH)COOH] + [CH,CH(OH)CH,])

3

2

87.6

105

85.5

54.5

36.8

Total concentration of the products on the basis of C content, μ mol dm" .

11.7

5.6

3.0

4.9

28.3

Electric charge passed during the electrolysis.

4.3

0.0

3.3

2.0

0.0

d

17.0

1.1

0.0

7.8

4.1

(%)

Volume of the catholyte after electrolysis.

21.4

99.0

1.0

2.7

2.8

(μΜ)

e

c

2.8

27.0

5.1

0.0

d

sc,

b

0.7

16.0

1.0

c

(Q

Q

Electrolysis Potential, -0.8 V vs A g / A g C l ; initial pH, 3.0; electrolysis time, 3 h; net area of the substrate, 19.3 cm .

0.0

9.8

3

(cm )

b

V

electrode

a

0.6

2.0

Acetic Lactic Acetone Formic

2-Pr

MeOH EtOH

( ml min )

1

Faradaic efficiency (%)

Flow rate

KCl

a

with a polymer composite/gas-diffusion

as a function of gas flow rate i n a solution of 0 . 5 M

reduction of C 0

Table III. Faradaic efficiencies for the products obtained i n the electrochemical

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

359 increased with an increase of the flow rate of C 0 , but became lower at the flow rate of 99.0 ml min' . At this flow rate, the electric charge passed during the electrolysis was also smaller. These results are probably attributed to the limited area of the liquid meniscus which was found to be important for the C 0 reduction on the mesh metal/gas-difi&ision electrode, because a part of the mesh electrode could not keep contact with the electrolyte under such a highflowrate. In the mediated reduction of C 0 , PB and ΡAn are in the reduced states as represented by reaction (8) and (9), respectively. 2

1

2

2

ffl

n

+

n

6

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

n

m [ F e ( C N ) ] + K + e- - » K Fe {Fe (CN) ] (PB) (Everitt's salt, ES) +

n

2

+

6

n

PAn /Fe L" + K + e-

-> PAn/Fe L7K

+

(8)

(9)

The reduction scheme of C 0 on the polymer-modified electrode (11) is schematically shown in Figure 7. In this scheme, C 0 is bifimctionally captured by ΡAn and the metal complex: the electrophilic carbon atom of C 0 is bound to the amino group of ΡAn, and the basic oxygen atom coordinates to the central metal of the complex. Electrons from the Pt substrate reach the PAn/PB interface where Yt is catalytically reduced to H on the zeolitic lattice of ES. The bifimctionally activated C 0 is hydrogenated by H . The reaction intermediates are stabilized in the nonaqueous atmosphere of conducting polymer, and various products including lactic acid and methanol are generated by further advanced hydrogénation. 2

2

2

a d s

2

ads

References 1. Halmann, M. Chemical Fixation of Carbon Dioxide; CRC Press: Boca Raton, FL, 1993. 2. Hori, Y.; Kikuchi, K.; Suzuki, S. Chem. Lett. 1985, 1695-1698. 3. Hori, Y.; Kikuchi, K.; Murata, Α.; Suzuki, S. Chem. Lett. 1986, 897898. 4. Cook, R. L.; MacDuff, R. C.; Sammells, A. F. J. Electrochem. Soc. 1987, 134, 1873-1874. 5. DeWulf, D. W.; Bard, A. J. Catal. Lett.. 1988, 1, 73-79. 6. DeWulf, D. W.; Jin, T.; Bard, A. J. J. Electrochem. Soc. 1989, 136, 1689-1691.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Figure 2

ofC0

7. Schematic

n

electrode.

of the electrochemical

on a Pt/PB/PAn-Fe L

representation

reduction

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch023

361 7. Hori, Y; Wakebe, H.; Tsukamoto, T.; Koga, Ο. Electrochim. Acta 1994, 39, 1833-1839. 8. Jermann, B.; Augustynski, J. Electrochim. Acta 1994, 39, 1891-1896. 9. Smith, B. D.; Irish, D. E . ; Kedzierzawski, P.; Augustynski, J. J. Electrochem. Soc. 1997, 144, 4288-4296. 10. Ogura,K.;Endo, N. J. Electrochem. Soc. 1999, 146, 3736-3740. 11. Ogura, K.; Endo, N.; Nakayama, M . J. Electrochem. Soc. 1998, 145, 3801-3809. 12. Reymond, J. P.; Mériaudeau, P.; Teichner, S. J. J. Catal. 1982, 75, 3948. 13. Sulymosi, F.; Erdohelyi, Α.; Kocsis, M . J. Chem. Soc., Faraday Trans. 1 1981, 77, 1003-1012.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 24

Use of Dense-Phase Carbon Dioxide in Catalysis 1

2

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

Bala Subramaniam and Daryle H. Busch 1

2

Departments of Chemical and Petroleum Engineering and Chemistry, University of Kansas, Lawrence, KS 66045

Starting in the early 90's, green chemistry research has been the focus of many investigations. This paper reviews heterogeneous and homogeneous catalysis in dense phase CO including CO -expanded solvents. The advantages of using dense phase CO include replacement of organic solvents with environmentally-benign CO , the enhanced miscibility of reactants such as O and H , alleviating interphase transport limitations, and the chemical inertness of CO . In addition, the physicochemical properties of CO -based reaction media can be pressure-tuned to obtain unique fluid properties (e.g., gas-like transport properties, liquid-like solvent power and heat capacities). The advantages of CO -based reaction media for optimizing catalyst activity and product selectivity are highlighted for a variety of reactions such as these: alkylation on solid acid catalysts; hydrogenation on supported noble metal catalysts; and a broad range of homogeneous oxidations with transition metal catalysts and dioxygen or organic peroxides as oxidant. The overview concludes with an assessment of the potential and outlook of CO -based reaction media in industrial chemicals processing. 2

2

2

2

2

2

2

2

2

2

Concerns about solvents have propelled research efforts aimed at developing environmentally-benign chemical processing techniques that either eliminate or significantly mitigate pollution at the source (J, 2, 3, 4, 5). A major research focus in chemical processing has been the replacement of traditional solvents with scC0 . Supercritical and near-critical dense carbon dioxide phases have been heralded as environmentally-benign solvents of great promise in many chemical applications, including the long used process for decaffeination of 2

364

© 2002 American Chemical Society

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

365 coffee beans, the chemical process industry, and, more recently, to replace the conventional organic solvents used in pharmaceutical processing and the cleaning of garments. The low toxicity and limited reactivity of CO2 make it suitable for use around foods and other consumer goods, and its low cost supports its use in very broad ranges of applications. Detailed reviews of reactions in sc media (6, 7, 8, 9, 10, //) are provided elsewhere. This overview is focused on applications of dense C 0 media in catalytic processing. Historically, emphasis has rested on scC0 , and the many advantages of that medium over traditional solvents are first discussed. Attention is then directed to the even greater advantages of C 0 expanded organic solvents, new media that were first introduced for oxidation reactions in our laboratories. These extremely variable solvent systems retain solubility characteristics of the two combined solvents and still enjoy the environmental advantages of more familiar dense C 0 media. Applications of dense phase C 0 in heterogeneous and homogeneous catalytic systems are first briefly reviewed. In homogeneous catalysis, the emphasis will be on oxidation systems that have been the subject of several years of collaborative research at the University of Kansas among researchers from the Departments of Chemical & Petroleum Engineering and Chemistry. Thus, homogeneous polymerization, hydrogénation reactions and phase-transfer catalysis are not reviewed in detail here. These are reviewed elsewhere [/2, 13]. Carbon dioxide is considered environmentally acceptable, non-toxic, relatively cheap (~5 cents/lb), non-flammable, inert toward oxidation and readily available. Supercritical reaction media, in general, have the potential to increase reaction rates, to enhance the selectivities of chemical reactions and to facilitate relatively easy separation of reactants, products, and catalysts after reaction. At ambient temperatures, the solubility of the much-favored terminal oxidant, atmospheric oxygen, in water and conventional organic solvents is limited. However, oxygen and carbon dioxide are miscible in all proportions under sc conditions. The increased concentration of reactants within an sc phase has the potential to accelerate reaction rates. In addition, the elimination of interphase (gas-liquid) mass transfer resistances may also increase reaction rates. The oneto-two-orders-of-magnitude higher intrinsic diffusivities of solutes in sc media compared to organic solvents could also accelerate reactions in the case of processes wherein the overall rate of reaction is limited by diffusion of reactants to form encounter pairs, which subsequently react to form products. The solubilities of solutes in sc fluids depend strongly on pressure and temperature in the vicinity of the critical point. Therefore, reactants, products and the catalyst can be separated by relatively simple pressure and/or temperature programming. Furthermore, if the phase behavior of the reaction mixture is such that the products fall out of the reaction mixture at the reaction Ρ and T, in situ product separation is achieved. However, in the case of C 0 , the dielectric constant remains virtually unaffected over a wide range of pressure. Thus, pressure-tuning effects in the sc region on conversion and selectivity may be investigated with 2

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

2

2

2

2

minimal change in solvent dielectric constant.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

366

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

Thermodynamic and Kinetic Aspects The principles developed for gas or liquid phase reactions may be applied to supercritical phase reactions as well. When the reaction medium density is gas­ like, the concepts developed for gas-phase reactions, such as kinetic theory of gases, may be applied. For liquid-like reaction mixtures, principles of liquidphase kinetics have been applied. Parameters such as the solvent's solubility parameter, dielectric constant or solvatochromic shift, routinely used to interpret liquid-phase reactions, have been employed to understand the effect of a given sc solvent on chemical reaction (6). In the vicinity of the critical point, sc reaction media admit sensitive pressure effects on reaction rate and equilibrium constants.

Pressure effects Reactions in sc media typically involve elevated pressures that could have either an enhancing or inhibiting effect on rate and equilibrium constants. Based on transition-state theory, the pressure dependence on the rate constant is given by the following equation:

31η*Λ

Δυ

dP J T ,

RT

x

where k is the mole-fraction based rate constant, Δ υ * is the activation volume (defined as the difference between the partial molar volume of the activated complex and sum of the partial molar volumes of the reactants), R is the universal gas constant, Ρ is the pressure and Τ is the absolute temperature. If the rate constant is related to a pressure-dependent measure of concentration (such as partial pressure), then the following equation is used: x

d\x\k\ dP

Δυ*

J τ,

RT

+

Κ

fl-n

d\nK

Ρ

d\nP

T

(2)

where k is the pressure-dependent rate constant, η is the molecularity of the reaction and κ is the isothermal compressibility factor. Equations (1) and (2) assume that the transition-state transmission coefficient is independent of pressure. In a similar manner, the effect of pressure on equilibrium constant is given by: c

τ

fd\nK\

Δυ

V dP J,,

RT

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

367 91η Κ dP

)

RT

(4)

^

where Δ υ is the reaction volume (defined as the difference between the partial molar volumes of reactants and products), K is the mole-fraction based equilibrium constant, K is the concentration-based equilibrium constant and v is the stoichiometric coefficient. The reported values of activation and reaction volumes in sc reaction media are up to two orders of magnitude greater those encountered in liquid phase reactions, implying that pressure effects on rate and equilibrium constants can be significant in sc phase reactions. The rather large Δν values in sc media are due to the fact that the partial molar volumes can assume large negative values near the critical point. Γ

x

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

c

t

Critical Phase Behavior of Fluid Mixtures Reliable knowledge of the phase border curves of reaction mixtures in pressure-temperature-composition (P-T-x) space is essential for identifying the appropriate operating pressure and temperature required for supercritical operation, to properly interpret conversion data, and to rationally develop postreaction product separation schemes. These phase border curves separate regions of different states of matter and can be two-phase liquid-liquid (LL) or liquidvapor ( L V ) boundaries, three-phase liquid-liquid-vapor ( L L V ) or solid-liquidvapor (SLV), or sometimes four-phase solid-solid-liquid-vapor ( L L S V ) boundaries. Any given system can admit one or more of these regions. Using three-dimensional P-T-x diagrams, lucid descriptions of the critical phase behavior exhibited by binary systems are available in the literature. For simple binary systems, where the critical L V locus of various binary mixtures is a continuous curve between the critical points of pure components, the L V phase border curve can usually be reliably predicted with an appropriate equation of state. In contrast, for binary systems that exhibit a region of liquid-liquid immiscibility (the so-called Type II system) or for ternary systems, predictive methods may not be reliable. The sc reaction mixtures surrounding a solid catalyst typically involve several components requiring experimental determination of the critical phase behavior. Detailed treatment of the classification of the types of phase behavior exhibited by fluid mixtures along with experimental and theoretical methods available for determining phase border curves may be found elsewhere (14).

Heterogeneous Catalytic Reactions The pressure-tunable density and transport properties of scC0 have been exploited in heterogeneous chemical reaction systems in various ways such as these: enhanced desorption and transport of heavy molecules (such as coke 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

368

precursors) in mesoporous catalysts alleviating pore-diffusion limitations and improving catalyst effectiveness; in situ removal of primary products stabilizing primary product selectivity; eliminating 0 or H solubility limitations in the liquid phase (and thereby eliminating interphase mass transfer resistances) in multiphase reaction systems; enhanced heat capacity ameliorating the problem of parametric sensitivity in exothermicfixed-bedreactors. Recent reviews (7, 8, 15) provide a comprehensive survey of the types of heterogeneous catalytic reactions investigated at supercritical conditions including alkylation, amination, cracking, disproportionation, esterification, hydrogénation, isomerization and oxidation. Some of these examples are described here to show how to systematically exploit sc media in heterogeneous catalysis.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

2

Alleviation of internal pore-diffusion resistances Pore-diffusion limitations in porous catalysts affect catalyst and/or product selectivity. Two instances are considered here. The first involves a parallel reaction network in which the desired product is produced in one reaction and "coke" precursors form via the second reaction. Accumulation of the coke precursors in the catalyst leads to catalyst "fouling" and causes a continuous decline in catalyst activity. The second example is a series reaction network of the type A -> Β -> C. Clearly, pore-diffusion limitations would hinder the removal of Β (the desired product)fromthe catalyst, favoring further reaction to the ultimate product, C. Examples are presented below to show how pressure tuning with sc media may be exploited to either desorb the coke precursors in situ (thereby stabilizing catalyst activity) or to desorb products, thereby enhancing intermediate product selectivity. Several applications of the in situ decoking concept have appeared in the literature. One such application involves stabilizing the activity of solid acid catalysts such as in alkylation reactions. As reviewed elsewhere (16), numerous efforts aimed at developing solid acid alkylation catalysts and solid acid-based isobutane-olefin alkylation processes have been reported for more than three decades. However, to-date, none of the solid alkylation catalysts has gained acceptance in industry for one or more of the following reasons: rapid catalyst deactivation due to coke formation, unacceptable product quality (i.e., low alkylate fraction) and thermal degradation of catalyst during the regeneration step. "Supercritical" alkylation, performed with excess isobutane (P = 36.5 bar; T = 135°C) above the critical temperature (T ) and critical pressure (P ) of isobutane, has been reported to slow down deactivation of solid acid catalysts (17, 18, 19). Fan et al. (18) carried out an investigation over Y-zeolite catalysts under liquid, gas, and near-critical phases. The use of sc reaction media provided extended alkylate activity compared to the liquid and gas phases, but the activity declined in all cases. However, at reaction temperatures exceeding 135°C, undesirable side reactions such as oligomerization and cracking dominate, resulting in unacceptable product quality. Clark and Subramaniam (20) c

c

c

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

c

369 employed C 0 (P = 71.8 bar; T = 31.1°C) as a diluent in the hydrocarbon feed to realize sc reaction mixtures with pressure-tunable properties in the 50-100°C range. The lower reaction temperatures favored alkylation reactions over oligomerization. The alkylate production attains a nearly steady value after a few hours on stream with .scC0 -based reaction media, indicating that scC0 -based reaction medium is capable of maintaining the activity of the strong acid sites. However, both the butene conversion (20%) and the alkylate selectivity (around 5%) are small suggesting the pore sizes and/or acidity in these catalysts are such that it is not possible to effectively balance the coke formation rate with the coke removal rate from the pores. Recent studies in our laboratory (21) show that enhanced alkylate selectivity can be obtained on a Si0 -supported Nafion catalyst (Engelhard SAC-13), which has an acidity comparable to sulfuric acid (used in conventional alkylate processing) and has relatively large pores when compared to zeolitic catalysts. Steady C alkylates production activity during experimental runs lasting up to two days was demonstrated during the alkylation of isobutene with 1-butene over silica-supported Nafion® catalyst particles suspended in a C0 -based supercritical reaction mixture in a slurry reactor. At a butene space velocity of 0.05 h' , 95°C (1.1 T ), molar feed I/O ratio of 5 with 70 mole% C 0 in feed, pressure-tuning studies revealed that while the butene conversion was relatively insensitive to pressure at 80% between 80 (-1.0 P ) and 167 bar (-2.2 P ), the C alkylates selectivity decreased fourfold from approximately from 30% at 80 bar to 7% at 167 bar (see Figure 1). The overall C selectivity decreased from approximately 74% to 30% in the same pressure range with heavier ( Q and higher) products being formed in denser supercritical reaction mixtures. Clearly, milder supercritical pressures provide the optimum combination of liquid-like densities and gas-like transport properties to desorb the C products and transport them out of the catalyst pores before they are transformed to heavier products. Glâser and Weitkamp (22) employed scC0 as a reaction medium to significantly reduce the deactivation rates of LaNaY-73 and H-mordenite zeolites during the isopropylation of naphthalene on these zeolites. These examples further demonstrate the ability of sc media for enhancing the stability of porous catalysts against deactivation by coke formation. The foregoing studies provide compelling evidence that C0 -based supercritical reaction mixtures in conjunction with properly designed solid acid catalysts (optimum pore structure, acidity and accessibility) offer an excellent opportunity for developing environmentally benign alternatives to conventional processes that employ mineral acids. 2

c

c

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

2

2

8

2

1

c

2

c

c

8

8

2

8

2

2

Eliminating inter-phase mass transfer resistances Conventional solid-catalyzed hydrogénation reactions are typically carried out in multiphase reactors involving the sparging of hydrogen through a slurry of the reactant and finely powdered catalyst particles. Typically, these reactions are fast and hence their rates are limited by the low solubility of hydrogen in the solvent. The other drawbacks of the conventional process include large reactor

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

370 volumes and costs associated with separating the product and the catalyst from the product mixture. Supercritical reaction media have been investigated for alleviating transport limitations in solid-catalyzed multiphase reactions such as in catalytic hydrogénations (23, 24, 25, 26). In addition to eliminating interphase mass transfer resistances, the enhanced extraction of heavy hydrocarbons and their facile transport from the catalyst pores by the near-critical reaction mixture alleviate internal pore diffusion limitations — as discussed in the preceding section. Consequently, increased catalyst effectiveness factors and enhanced primary product selectivities (in the case of a series reaction network) are observed when pressure-tuning in the near critical region. The ensuing enhanced reaction rates reduce holdup of hazardous reactants making the process inherently safer. Additionally, due to its liquid-like heat capacity, scC 0 -based hydrogénation offers the possibility of performing the reactions in a fixed-bed reactor with controlled temperature rise in the bed. Further, product separation from C 0 is achieved by depressurization of the reactor effluent stream. By solubilizing the reactants (organic substrate and hydrogen) in a single, environmentally benign sc solvent such as C 0 , or propane, it has been shown that selective hydrogénation of vegetable oils and fats (to cis- fatty acid in preference to trans- fatty acid) can be performed on supported Ni catalysts at significantly high rates (about 400 times faster) relative to conventional slurry phase operation (23). Employing scC0 as a reaction medium, a wide variety of organic functional groups (alkene, cyclic alkenes and alkanes, aldehydes, ketones, etc) can be hydrogenated with good throughput on polysiloxane-supported noble metal catalysts such as Pd and Pt (27). Supercritical propane was preferred in the case of nitrogen-containing compounds to avoid the formation of carbamic acid salts by the reactions of amine groups with C 0 . Employing dense C 0 as the solvent medium, Bertucco et al. (30) systematically investigated the hydrogénation of unsaturated ketones by a supported Pd catalyst in a stirred reactor to minimize transport gradients. Krôcher et al. (28) demonstrated the synthesis of N, N- dimethyl formamide and methyl formate from a homogenous mixture of C 0 and H , using heterogeneous silica-based hybrid-type Ru catalysts. Thus, C 0 functions as both a reactant and a solvent. The turnover frequencies with the heterogenized catalyst were twofold higher than the homogeneous counterpart. During hydrogénation of cylcohexene on supported Pd catalysts using scC0 as solvent, significant axial temperature gradients exceeding 100°C were reported by Hitzler et al. (27). Another important issue during hydrogénation in scC0 is that of catalyst deactivation. Minder et al. (29) investigated the batch hydrogénation of ethyl pyruvate to (R)-ethyl lactate over a modified Pt/Al 0 catalyst in sc reaction media. When using sc C 0 as the reaction medium, they report catalyst deactivation, presumably as a result of Pt poisoning by CO formed from the reverse water-gas shift reaction between C 0 and H . In contrast, none 2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

3

371 of the continuous catalytic hydrogénation studies in scC0 (23, 24, 30) report catalyst deactivation. Recently, we revisited the hydrogénation of cyclohexene in scC0 on a supported Pd catalyst to investigate temperature control and catalyst deactivation issues (31). At near-critical experimental conditions (136 bar and 70°C), the maximum temperature rise observed was 12°C, which is less than the adiabatic temperature rise at total conversion. In the presence of approximately 180 ppm peroxides in the feed, there is a gradual but steady deactivation of the catalyst (roughly 2%/h) during the 18 h run. The spent catalyst at the end of the run showed significant surface area and pore volume losses. In contrast, when the peroxide content in the feed is mitigated to 96% selectivity. Wu and coworkers (50) demonstrated the catalysis by Fe(PFTPP)Cl of the aerobic oxidation of cyclohexane in the presence of acetaldehyde to cyclohexanol and cyclohexanone. Typical reaction conditions were 1 h reaction time, 70°C, 1.0 mmol of cyclohexane, 0.25 mmol of acetaldehyde, 0.5 μιτιοί of catalyst, 1.0 mmol of 0 and 90 bar, giving a total yield of 5% of cyclohexanol and cyclohexanone. We studied the homogeneous oxidation catalysis of CH Re0 (WiTO) and olefin epoxidation in scC0 (10). We chose /-butyl hydroperoxide and completely homogeneous reactions were run for 48 h in a 60 mL Jurgeson cell at 2

!

3

!

2

6

4

2

2

2

3

2

2

2

3

3

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

Figure 3. Catalyst/Ligand structures.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

376 approximately 83 bar and 50°C. The terminal oxidant was delivered as a /BuOOH/i-BuOH mixture that served as a co-solvent, substantially increasing the solubility of the catalyst. We identified the products and determined the yields from catalytic oxidation of cyclohexene, styrene, and cis- and /nms-stilbene. Whereas the oxidation of cylohexene yielded the oxide and its hydrolysis product, the corresponding diol, the rest of the substrates also underwent C-C bond cleavage and yielded benzaldehyde, in addition to their epoxides. Conversion for cyclohexene was 64% and those for styrene and the stilbene isomers were similar. These preliminary results indicate that scC0 is a viable solvent for this catalyst system. Similar results were obtained with Jacobson's catalyst, Mn(salen*) (see Figure 3), which is sufficiently soluble in scC0 to catalyze the epoxidation of a range of olefins using /-BuOOH in ί-BuOH co-solvent as terminal oxidant. All reactions were run with 50 μΜ of catalyst and 83 bar total pressure at 50 °C and 70 % w/w in f-BuOOH/f-BuOH. As observed by Kolis (47) and in our work with CH Re0 , epoxidation is accompanied by C=C cleavage in the case of olefins having phenyl groups as substituents. Our results are summarized as follows (substrate is followed by % yield of epoxide and then by % C=C cleavage: cyclohexene, 27% (epoxide 15% plus diol 12%, no cleavage); styrene, 17%, 23%; c/s-stilbene, 12%, 19%; /ra^s-stilbene, 16%, 20%. Kochi and coworkers have found similar results using the manganese complex of a different substituted salen, Mn{(5,5-N0 ) salen}, and iodosobenzene as the terminal oxidant in CH CN as the solvent (51). Kochi's results are: cyclohexene, 56%, no cleavage; styrene, 37%, 10%; cw-stilbene, 35%, 7%; /ra/w-stilbene, 40%, 6%.; showing considerable more selectivity toward epoxidation over C=C bond cleavage. While showing no conversion or selectivity advantage, the experiments with Mn(salen*) in scC0 indicate the viability of dense C 0 media for reactions involving this important catalyst, thereby retaining the possibility of an environmental sustainability advantage. Oxygen is completely miscible with ,scC0 media, but its solubility in most solvents is only a few millimoles per liter. This strongly suggests that scC0 and other C 0 derived media should enjoy advantages in reactions using 0 as terminal oxidant. Proceeding on this basis, we have investigated well known oxidation reactions catalyzed by cobalt(II) dioxygen carriers, the oxidation of phenols to quinones. Further, the catalysts chosen were the inexpensive parent of the family, Co(salen) itself, and a second member of the family selected because its ligand (from Jacobson's catalyst) imparts high solubility to metal complexes in nonpolar solvents and because it too is readily available. The catalytic oxidations of 2,6-di-tert-butylphenol, DTBP, and 3,5-di-ter/butylphenol, 35-DTBP, in scC0 were studied in the presence of a large excess of 0 , 207 bar total pressure and a reaction temperature of 70°C (52). Under the experimental conditions, DTBP is converted to a mixture of just two products, 2,6-di-/ert-butyl-l,4-benzoquinone, DTBQ, and the related product of radical

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

2

3

3

2

2

3

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

377

coupling, 3,5,3\5Metra-teri-butyl-4,4'-diphenoquinone, TTDBQ. It should also be recognized that DTBQ and related materials are of interest in medicinal chemistry (53). Conversion and selectivity were studied as functions of temperature, pressure, and concentrations of catalyst, substrate, and terminal oxidant, 0 , and the results provide substantial mechanistic elucidation. The contrasting temperature dependence of the conversion to products, versus that of selectivity between them, is most revealing. Conversion increased steadily with temperature as expected for an initial oxidation step involving a substantial activation energy; i.e., the formation of the phenoxy radical. In contrast, selectivity between the pathways producing the competing products is meaningfully temperature independent. This is consistent with the conclusion that the final product mixture is determined by competition between two parallel reactions, both of which are essentially radical-radical coupling reactions with vanishingly small activation energies. Thus, the very essence of the proposed mechanism for these long known reactions is strongly supported. The dependences of conversion and selectivity on oxygen concentration are equally revealing. Both parameters obey saturation models, a result that links both rate determining and product determining processes to the dioxygen complex of Co(salen). Since both conversion and selectivity become independent of oxygen concentration at high 0 concentrations, it is clear that free 0 is not involved in the initiation step that determines conversion or either of the parallel reactions that determine the relative yields of the products. Whereas there was no doubt that the initiation step involved the oxygen complex of Co(salen), this provided the first proof that the oxygen complex is responsible for capture of the phenoxy radical along the pathway that leads to DTBQ. It is also implicit in this result that no direct reaction occurs between the electrophilic phenoxy radical and 0 . The observed behavior provides strong support for the previously proposed mechanism for these reactions (54, 55).

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

2

2

2

Oxidations in CC^-Expanded Solvent Media Significant limitations accompany the well-known advantages of s o C 0 for catalytic oxidations, including low reaction rates, high process pressures (on the order of hundreds of bars), and the limited number of transition metal catalysts that are sufficiently soluble in C 0 without substantial structural modification. In processes demonstrated recently by us, conventional solvents are substantially, but not totally replaced by dense C 0 . The retained organic solvent preserves some of its desirable properties in the mixture, e.g., dielectric constant, solvation and coordination properties. The solution volume is multiplied by the C 0 expansion without separation from the reaction mixture of the substrate, catalyst, or terminal oxidant from the reaction mixture, creating a completely homogeneous C0 -expanded reaction mixture. The total pressure is ~ 50-90 bars 2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

378 (much lower than typical pressures for sc-C0 ) and the C 0 mole fraction in the solvent is typically between 65 and 80%, a substantial reduction in organic solvent usage. Further, our measurements (56) have shown that the preferred terminal oxidant dioxygen is at least one or two orders of magnitude more soluble in these media than in the neat organic solvents. The examples that follow show striking advantages, including enhanced oxidation, improved selectivities, and increased operational safety. Adding C 0 to solvents having significant solubilities in C 0 "expands" the solvent while decreasing the solubilities of the dissolved solutes in the mixture (57, 58, 59). This concept has also been used in separation schemes (13). We investigated the effect of isothermal C 0 addition to a solution of C H C N containing dissolved catalyst [iV,yV-Bis(3,5-di-^-butylsalicylidene)l,2cyclohexanediaminato(2-)] cobalt(II) {i.e., Co(Salen*)}. The volume of the homogeneous liquid phase expands rapidly as C 0 is added. The catalyst solubility in the expanded solvent decreases continuously with the expansion and the catalyst precipitates out at a certain level of expansion, termed as the "maximum homogeneous expansion limit" (V/V ) x. The expansion factor is defined as the ratio of C0 -expanded solvent volume (V) to the initial solvent volume (V = 5 mL). For a given catalyst/solvent combination, the maximum homogeneous expansion limit decreases with increasing catalyst concentration and increases at higher temperatures. For the catalyst concentration employed in our reaction studies (0.4 mg/mL CH CN), the maximum homogeneous expansion limit is roughly 5 at 50°C. This implies the possibility of replacing up to 80 vol.% of the organic solvent by C 0 while maintaining the catalyst in solution. It is noteworthy that the total pressures at the maximum homogenous expansion limit are tens of bars, as compared to hundreds of bars for scC0 2

2

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

2

2

3

2

0

ma

2

0

3

2

2

Another advantage of C0 -expanded reaction media is the enhanced solubility of 0 . The measured 0 mole fraction in C0 -expanded CH CN (V/V = 2) is about two orders of magnitude higher than the 0 solubility in neat CH CN ((5xl0" molefractionat 25°C and 1 bar [60]), and is of the same order of magnitude as that observed in liquid C 0 (61). For comparison purposes, the homogeneous catalytic 0 -oxidation of 2,6-ditert-butylphenol, DTBP, by Co(Salen*) was studied in sc-C0 (scheme given in Figure 4), in C0 -expanded organic solvents (V/V = 2), and in neat organic solvents. The common conditions are as follows: Catalyst amount = 0.41 mmol; catalyst:substrate:0 mole ratio = 1:40:80; volume of methyl imidizole = 2 \iL. As seen from Figure 5, the turnover frequency (TOF), defined as the moles of DTBP converted per mole of catalyst per hour, in the C0 -expanded CH CN (P = 60-90 bars) is between one and two orders of magnitude greater than in scC0 (P = 207 bar). The observed selectivity toward DTBQ (80-88%) is comparable in scC0 and C0 -expanded CH CN and no CH CN oxidation was detected in either case. The TOF and DTBQ selectivity are lower in neat C H C N (0 bubbling, 28°C, 1 bar). Neat solvents can form explosive gaseous mixtures at higher temperatures and hence were avoided. The high TOFs observed in C 0 2

2

2

2

3

0

2

4

3

2

2

2

2

0

2

2

3

2

2

2

3

3

3

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

379 expanded CH CN are attributed to the ability of the polar solvent to stabilize the polar transition state, thereby lowering the activation energy and increasing the reaction rate. The rather sensitive temperature dependence of TOFs in the C 0 expanded phase (more than 10-fold increase for the range 25-80°C) compared to scC0 (2-fold increase over the 50-80°C range) supports this mechanistic hypothesis. The inexpensive and readily available Co(salen) catalyst is insoluble in scC 0 . In sharp contrast, Co(salen) shows remarkable activity in C 0 expanded CH C1M. These results show that C0 -expanded solvents complement scC0 as reaction media by broadening the range of conventional catalyst+solvent combinations with which homogeneous oxidations by 0 can be performed. Taking advantage of the solubility advantages, the oxidation of cyclohexene was investigated with a non-fluorinated iron porphyrin catalyst, (5,10,15,20tetraphenyi-21//,23//-porphyrinato)iron(HI) chloride, Fe(TPP)Cl, in addition to a fluorinated catalyst (5,10,15,20-tetrakis(pentafluoropheny l)-21//,23//porphyrinato)iron(III) chloride, Fe(PFTPP)Cl (56). In both cases, molecular oxygen was used as the terminal oxidant. While Fe(TPP)Cl is insoluble and displays little activity in scC0 , it displays remarkably high activity in C 0 expanded CH CN. GC/MS analysis revealed five products: cyclohexene oxide, 1,2-cyclohexane-diol, 2-cyclohexene-l-ol, 2-cyclohexene-l-one, and 4-hydroxy2-cyclohexene-l-one. While the first two products are those of oxygen atom transfer, the rest are formed by a mechanism that involves allylic H-abstraction. The 1,2-cyclohexanediol is formed by hydration of cyclohexene oxide and, therefore, the epoxidation selectivity is based on the yields of both cyclohexene oxide and cyclohexanediol. Conversion histories of cyclohexene oxidation with Fe(TPP)Cl at 50°C in C0 -expanded CH CN (V/V = 2) and in neat CH CN reveal an "induction" period, which presumably involves the buildup of radicals to a critical concentration. Remarkably, the induction period in C0 -expanded CH CN is only 4 hours compared to nearly 16 hours in neat CH CN. We attribute this reduced induction period to the enhanced 0 solubility in C 0 expanded CH CN, assuming that the free radical initiation step involves molecular oxygen. As shown in Table 1, the cyclohexene conversion obtained with the fluorinated catalyst Fe(PFTPP)Cl in the C0 -expanded solvent at 90 bar and 80°C is 41% which is nearly 7-fold greater than that reported by Birnbaum et al. (45) in scC0 with the identical catalyst at the same temperature, but 345 bar. Also, an approximately 1.5 fold higher epoxidation selectivity over scC0 is obtained in C0 -expanded CH CN. It is interesting to note that the nonfluorinated catalyst, Fe(TPP)Cl, which is insoluble in scC0 , displayed a higher TOF for cyclohexene oxidation, fully 10-fold greater than the result in scC0 , with comparable epoxidation selectivity. Cyclohexene conversion and product distribution are sensitive to the C 0 fraction in the reaction medium. Conversion increases from 24% in neat CH CN to a maximum of 31%, at 2-fold expansion, decreasing upon further C 0 expansion. Epoxidation selectivity showed a similar trend. We attribute this behavior to the lower 0 solubility in neat organic solvent and the progressively 3

2

2

2

2

3

2

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

2

2

2

3

2

3

G

3

2

3

3

2

2

3

2

2

2

2

3

2

2

2

3

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

380

0 + LCo

^

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

OH

LCoOO.

Ο

Ο

Ο Figure 4. Scheme for the oxidation of 2,6-di-tert-butyl phenol by Schiff base cobalt catalysts.

Table 1. Cyclohexene oxidation by iron porphyrin complexes in different reaction media p, bar

Reaction medium

X,%

S,%

Catalyst

Ref

90

C0 -expanded C H C N

55

19.2

Fe(TPP)Cl

(56)

90

C0 -expanded C H C N

40.9

28.3

Fe(PFTPP)Cl

(56)

345

scC0

5.5

19

Fe(PFTPP)Cl

(45)

2

3

2

3

2

p: pressure; X : conversion; S: epoxidation selectivity. Conditions: 80°C; 4 h; 10 mL reactor; 2.8 μπιοί of catalyst; catalyst:0 :cyclohexene mole ratio = 1:1500:2000; (V/V ) = 2 (for C 0 expanded CH CN). 2

0

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

381

100 Co(salen*) in expanded CH CN I I ^ (50 - 90 bar) •

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

3

Co(salerf) in neat CH CN (1 bar) 3

0.1 20

40 60 80 Reaction Temperature, °C

Figure 5. Turnover frequencies Co(Salen*)

and

100

(TOFs) for 2,6-di-tert-butyl phenol oxidation

with

Co(salen).

diminishing solvent polarity with increasing fraction of C 0 . This indicates that a continuum of reaction media with different properties may be realized by varying the C0 /solvent ratio, and that the ratio may be optimized for a given process through a combination of solubility effects and solvent properties. 2

2

Concluding Remarks The pressure-tunability of the density and transport properties of near-critical fluids can be exploited for achieving unique combinations of fluid properties (i.e., liquid-like densities yet significantly better transport properties compared to liquids). These properties are better suited than either gas phase or liquid phase reaction media to alleviate transport limitations in heterogeneous fluidsolid catalytic reactions. To attain pressure-tunable reaction mixtures, the reaction temperature must be in the vicinity of the critical temperature of the reaction mixture (1.05 - 1.2 T in general). To satisfy this criterion, the F of the reaction mixture may have to be suitably modified by adding an inert cosolvent such as carbon dioxide. For this purpose, reliable knowledge of the critical phase behavior of the reaction mixture including various cosolvent candidates is essential. The operating pressure that provides the optimum fluid properties for maximizing a given performance criterion (catalyst activity, effectiveness or selectivity) is determined by pressure-tuning studies. Higher-than-optimum pressure could introduce transport limitations that diminish catalyst effectiveness factor while lower-than-optimum pressure could limit the desorption of the heavy hydrocarbons, adversely affecting catalyst activity and product selectivity. The c

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

c

382 several examples provided in this chapter demonstrate how the pressure-tunable density and transport properties of supercritical reaction media have been exploited in heterogeneous catalysis such as: (a) enhanced desorption and transport of heavy molecules (such as coke precursors) in mesoporous catalysts alleviating pore-diffusion limitations and improving catalyst effectiveness; (b) in situ removal of primary products stabilizing primary product selectivity; and (c) eliminating 0 or H solubility limitations in the liquid phase (and thereby eliminating interphase mass transfer resistances) in multiphase reaction systems; and (d) enhanced heat capacity ameliorating the problem of parametric sensitivity in exothermic fixed-bed reactors. One or more of these advantages have been demonstrated for several classes of heterogeneous (i.e., solid-catalyzed) reactions such as alkylations, aminations, hydroformylations and hydrogénations, spanning a wide spectrum of chemical process industries including petrochemicals, fine chemicals, food, agricultural chemicals and pharmaceuticals. The use of sc media in these examples is more than merely as a replacement for organic solvents. In virtually every case, the sc reaction medium represents an enabling tool used to manipulate or control such factors such as catalyst stability, product selectivity and temperature rise in the reactor. Conventional reaction engineering theory provides us the approaches and tools for the modeling and rational analysis of sc phase reactors. It follows from the many investigations reviewed herein that while the use of scC0 in homogeneous catalytic oxidation has certain advantages over conventional solvents (such as total solvent replacement with an environmentallybenign solvent, complete 0 miscibility in the reaction mixture, resistance to oxidation), a major drawback is the high pressures (on the order of hundreds of bars) required to ensure adequate solubility of many transition metal catalysts in C 0 . Fluorocarbon additives are known to enhance the solubility of transition metal complexes in s c C 0 at moderate pressures; however, the fluorocarbon compounds are expensive. Recently, poly (ether-carbonate) copolymers that readily dissolve in C 0 at lower pressures have been developed (62). Such research efforts aimed at developing low-cost, fluorine-free alternatives are expected to continue. Another limitation of s c C 0 is the lack of tunability of the dielectric constant and other properties of the reaction medium. By adding a suitable solvent that is miscible with dense C 0 (C0 -expanded solvent), such properties as the dielectric constant may be readily varied with mixture composition while maintaining solubility of the substrate, oxidant and catalyst in the expanded phase. There are a number of traditional solvents, such as CH C1 , CH CN, DMF, acetone, methanol, ethanol, tert-butmo\ and isopropanol, that can be expanded with dense C 0 . The C0 -expanded solvent mixtures represent a continuum of reaction media with different physicochemical properties that may be exploited for optimizing conversion and selectivity in well-known catalytic oxidation systems as discussed herein. Further, the low process pressures and the ability to separate and recycle the solvent make homogenous catalytic oxidation in C0 -expanded solvents environmentally benign and more economical compared to scC0 -based oxidation.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

2

2

2

2

2

2

2

2

2

2

3

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

383 Another kind of environmentally-benign solvent that is receiving increased attention is room temperature ionic liquids such as chloroaluminate(III) ionic liquids, specifically 1-butylpyridinium chloride-aluminum(IH) chloride and 1ethyl-3-methylimidazolium chloride-aluminum(III) chloride. Ionic liquids are extremely good solvents for a wide range of inorganic, organic and polymeric materials. Recently, it was shown that by sufficiently expanding ionic liquids with scC0 , hydrocarbon components can be selectively extracted from the reaction mixture (63). This suggests that controlled expansion of room temperature ionic liquids with CO2 should be possible such that the hydrocarbon substrate and catalyst miscibilities in the reaction mixture are maintained while enhancing the solubilities of the such oxidants as dioxygen therein. Post reaction, the hydrocarbon components and the catalyst may be fractionated from the reaction mixture by stepwise addition of dense C 0 . Catalytic oxidations in C0 -expanded ionic liquids thus offer the exciting possibility of developing novel integrated environmentally-benign reaction and separation schemes. It should be clear that the advantages of C0 -expanded phases can also be realized with heterogeneous catalytic systems in which the solid catalyst particles are exposed to the C0 -expanded reaction mixture. Another area that shows much promise for exploitation in environmentally-benign oxidations is the use of biphasic C0 /water emulsions (64). Water soluble catalysts are sequestered in highly dispersed water phase while the substrate and oxidant (such as dioxygen) are dissolved in the continuous scC0 phase. The outlook for industrial applications of supercritical reaction media in heterogeneous fluid-solid catalysis is promising. Economic factors that dictate the viability of supercritical processes include the capital costs associated with the high-pressure equipment and operating costs associated with the compression of the supercritical media for recycling purposes. Many examples discussed in this chapter show that the optimum pressure lies close to the critical pressure of the reaction mixture. Clearly, the critical pressure of the solvent medium and the extent of dilution of the reactants with the solvent medium are important parameters that would dictate the process economics. The announcements (65) of a development plant for producing fluoropolymers in supercritical carbon dioxide and of a supercritical phase hydrogénation plant are strong indicators that industries are seriously considering sc processes for chemical processes. The challenge will be in demonstrating processes that simultaneously display the following attributes: selective, environmentally-benign, stable and economical.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

2

2

2

2

2

2

2

Acknowledgments This work was supported in part by the National Science Foundation (CHE9815321; CTS-9816969) and the Environmental Protection Agency (R82703401-0).

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

384

References 1.

Collins, T. J., Green chemistry in Macmillan Encyclopaedia of Chemistry, Macmillan, Inc., New York (1997).

2.

Anastas, P. T. & Williamson, T. C., Green chemistry: an overview, in Green Chemistry: Designing Chemistry for the Environment, American Chemical Society Symposium Series No. 626 (ed. Anastas, P.T. & Williamson, T.C.), American Chemical Society, Washington, D.C., 1996, pp 1-17. Anastas, P.T. & Warner, J.C., Green chemistry: Theory and Practice, Oxford University Press, Oxford (1998). Clark, J. H., Green Chemistry, 1999, 1, 1-8. Several articles in Anastas, P. T., Heine, L. G. and Williamson, T. C. (Eds.), in Green Engineering, American Chemical Society Symposium Series 766, 2000. Savage, P.E. , Gopalan, G., Mizan, T. I., Martino, C.J. and Brock, E.E., AIChE J., 1995, 41, 1723-1778. Savage, P. E. in Handbook of Heterogeneous Catalysis, eds. Ertl, G., Knözinger, H. and Weitkamp, J. Vol. 3, Wiley-VCH, Weinheim, 1997, pp 1339-1344. Baiker, A., Chem. Rev., 1999, 99, 453-473. Hutchenson, K. W., in Supercritical Fluid Technology in Materials Science and Engineering, ed. Y-P. Sun, (in press) Marcel Dekker, NY. Musie, G., Wei, M., Subramaniam, B. and Busch, D. H. Coord. Chem. Reviews., in press. Chemical Synthesis Using Supercritical Fluids, eds. Jessop, P. G., Leitner, W., Wiley-VCH, Weinheim, 1999. Several chapters in Chem. Rev., 1999, 99. Eckert, C. Α., Bush, D., Brown, J. S. and Liotta, C. L. Ind. Eng. Chem. Res., 2000, 39, 4615-4621. McHugh, M. A. and Krukonis, V. J. Supercritical Fluid Extraction, 2 Ed., Butterworth-Heinemann, Boston, 1994. Subramaniam, B., Appl. Catal A Gen., 2001, 212, 199-213. Weitkamp, J. and Traa, Y. Catal. Today, 1999, 49, 193-199. Hussain, Α., U.S. Pat. #5,304,698 (1994) to Mobil Oil Corporation. Fan, L., Nakamura, I., Ishida, S. and Fujimoto, K.L. Ind. Eng. Chem. Res., 1997, 36, 1458-1463. Funamoto, G., Tamura, S., Segawa, K., Wan, K. and Davis, M.E., Res. Chem. Intermed., 1998, 24, 449-459. Clark, M. C. and Subramaniam, B. Ind. Eng. Chem. Res., 1998, 37, 12431250. Lyon, C. J., Subramaniam, B. and Pereira, C. J., 2001, Stud. Surf. Sci. (in press). Gläser, R. and Weitkamp, J., in Proc. 12 Int. Zeolite Conf., Vol. 2, ed. M. Treacy, M. J., Mat. Res. Soc., Warendale, PA, 1999, pp 1447-1454.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

3. 4. 5.

6. 7.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

nd

th

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

385 23. Härröd, M. and Møller, P. in High Pressure Chem. Engng, Proc. Tech. Proc. 12, eds., Ph. Rudolf von Rohr, Ch. Trepp, Elsevier, Amsterdam, 1996, pp 4351. 24. Hitzler, M. G. and Poliakoff, M. Chem. Commun., 1997, 1667-1668. 25. Bertucco, Α., Canu, P., Devetta, L. and Zwahlen, Α., lnd. Eng. Chem. Res., 1997, 36, 2626-2633. 26. Tacke, T., Wieland, S. and Panster, P. in High Pressure Chem. Engng., Proc. Tech. Proc. 12, eds., Ph. Rudolf von Rohr, Ch. Trepp, Elsevier, Amsterdam, 1996, pp 17. 27. Hitzler, M. G., Smail, F. R., Ross, S. K. and Poliakoff, M. Org. Process Res. Dev., 1998, 2, 137-146. 28. Kröcher, Ο., Köppel, R. Α., Fröba, M. and Baiker, Α., J. Catal., 1998, 178, 284-298. 29. Minder, B., Mallat, T., Pickel, K. H., Steiner, K. and Baiker, A. Catal. Lett., 1995, 34, 1-9. 30. Devetta, L., Giovanzana, Α., Canu, P., Bertucco, A. and Minder, B. J. Catal. Today, 1999, 48, 337-345. 31. Arunajatesan, V., Subramaniam, B., Hutchenson, K. W. and Herkes, F. E. Chem. Eng. Sci., 2001, 56/4, 1363-1369. 32. Clark; M. C. and Subramaniam, B. Chem.Eng.Sci., 1996, 51, 2369-2377. 33. Dooley, K.M. and Knopf, F.C., lnd. Eng. Chem. Res., 1987, 26, 1910-1916. 34. Suppes, G.J., Occhiogrosso, R.N. and McHugh, M.A., lnd. Eng. Chem. Res., 1989, 28, 1152-1156. 35. Zhou, L. and Akgerman, Α., lnd. Eng. Chem. Res., 1995, 34, 1588-1595. 36. Wang, C.-T. and Willey, R.J. J. Non-Cryst. Solids, 1998, 225, 173-177. 37. Oakes, R.S., Clifford, A.A., Bartle, K.D., Pett, M.T. and Rayner, C.M., Chem. Commun., 1999, 247-248. 38. Fan,L.,Watanabe, T. and Fujimoto, K. Appl. Catal. A: General, 1997, 158, L41-L46. 39. Gaffney, M . and Sofranko, J. A. Preprints, American Chemical Society Division of Petroleum Chemistry, 1992, 37, 1273-1279. 40. Meehan, N. J., Sandee, A. J., Reek, J. N. H., Kamer, P. C. J., van Leeuwen, P. W. N. M. and Poliakoff, M. Chem. Commun., 2000, 1497-1498. 41. Vieville, C., Mouloungui, Z. and Gaset, Α., lnd. Eng. Chem. Res., 1993, 32, 1497-1498. 42. Lin, Y-H., Brauer, R. D., Laintz, K. E., and Wai, C. M. Anal. Chem., 1992, 65, 2549-2551. 43. Kainz, S., Koch, D., Baumann and Leitner, W., Angew. Chem. Int. Ed. Engl., 1997, 36, 1628-1630. 44. Morgenstern, D. Α., LeLacheur, R. M., Morita, D. K., Borkowsky, S. L., Feng, S., Brown, G. H., Luan, L., Gross, M. F., Burk, M. J., and Tumas, W., in Anastas, P. T. and Williamson, T. C., Eds., Green Chemistry: Designing Chemistry for the Environment, ACS Symposium Series 626, Washington, D.C., 1996, pp. 132-151. 45. Birnbaum, E. R., Le Lacheur, R. M., Horton, A. C. & Tumas, W., J. Mol. Catal., 1999, A 139, 11-24.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

386 46. Pesiri, D. R., Morita, D. K., Glaze, W. and Tumas, W., , Chem Commum,, 1998, 1015-1016. 47. Hass, G.R. and Kolis, J.W., Tetrahedron Lett. 1998, 39, 5923-5926. 48. Haas, G. R. and Kolis, J. W, Organometallics, 1998, 17, 4454-4460.

Downloaded by UNIV OF GUELPH LIBRARY on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch024

49. Jia, L., Jiang, H. and Li, J., Chem. Commun., 1999, 985-986. 50. Wu, X.-W., Oshima, Y. and Koda, S., Chem. Lett., 1997, 1045-. 51. Srinivasan, K., Michand, P. and Kochi, J. K., J. Am. Chem. Soc., 1986, 108 2309-2320. 52. Musie, G. T., Wei, M., Subramaniam, B. and Busch, D. H., Inorg. Chem., 2001, 40(14), 3336-3341.. 53. Kolesnik, I. G., Zhizhina, E. G and Matveev, K. I., J. Mol. Catal. 2000, 153, 147-154. 54. Zombek, Α., Drago, R. S., Corden, Β. B. and Gaul, J. H., J. Am. Chem. Soc., 1981, 103, 7580-758 . 55. Deng, Y. and Busch, D. H., Inorg. Chem., 1995, 34, 6380-6386. 56. Wei, M., Musie, G. T., Busch, D. H. and B. Subramaniam (submitted for publication). 57. Gallagher, P. M., Coffey, M. P., Krukonis, V. J. and Klasutis, N, in Amer. Chem. Sym. Ser., No. 406, American Chemical Society, Washington, D.C., (1989). 58. De la Fuente Badilla, J.C., Peters, C. J. and Arons, de Swaan, J., J. Supercritical Fluids, 2000, 17, 1-13. 59. Tai, C. Y and Cheng, C-S., AIChE J, 1998, 44, 989-992. 60. Hildebrand, J. H., Prausnitz, J. M. and Scott, R. L., Regular and Related Solutions: the Solubility of Gases, Liquids, and Solids, New York, Cincinnati, Toronto, London, Melbourne(1970). 61. Battino R. et al. (eds), Solubility Data Series: Volume 7, Oxygen and Ozone, Pergamon Press, Oxford, New York, Toronto, Sydney, Paris, Frankfurt (1981). 62. Sarbu, T., Styranec, T., Beckman, E. J., Nature, 2000, 405, 165-168. 63. Blanchard, L. Α., Hancu, D., Beckman, E. J., & Brennecke, J. F., Nature, 1999, 399, 28-29. 64. Jacobson, G.B., Lee, Jr., C.T., Johnston, K. P. and Tumas, W., J. Am. Chem. Soc. 1999, 121, 1902-1903. 65 . M. McCoy, Chem. Eng. News, 14 June 1999, 11.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Chapter 25

The Possibility of Coupling Supercritical Extractions in Petroleum Industry with CO Conversion Plants to Lessen Investments Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

2

Marco Antonio G. Teixeira and Maria Luísa A. Gonçalves Petrobras Research and Development Center, Division of Chemistry, Rio de Janeiro 21949-900, Brazil

The investment and the operational costs are strong limiting factors for the CO conversion processes, as they are for supercritical extraction implementation. A process solution to that is proposed, i n which supercritical extraction with CO , that employs high temperature and pressure, is used to yield valuable products before admittance into a CO conversion unit, thus sharing the energy costs with it. The extracted material is recovered by depressurization of the system, and the resulting CO stream can be obtained in conditions for conversion. A n application from petroleum industry is discussed, the extraction of valuable paraffins that also produces value-added paraffin-free asphalt. The temperature and pressure at which the CO stream can be obtained after liberation of extracted paraffins are adequate as feedstock preparation for most of the conversion propositions in literature. 2

2

2

2

2

C 0 mitigation has been the matter of several researches in the decade of the 90's, because of consequences of its liberation to the atmosphere to the global weather. Several scientific and technological meetings, like the 2

© 2002 American Chemical Society In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

387

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

388 1992, 1994 and 1996 International Conferences on Carbon Dioxide Removal (ICCDR) and the 1998 and 2000 International Conferences on Greenhouse Gas Control Technologies (GHGT), have discussed the mitigation possibilities, among which the conversion to products of commercial value, the main subject of the present publication, is a promising option. The technological development of process alternatives is a task that has been carried out in a very successful way by several researchers, many of them with contributions to this issue. However, largescale applications require industrial plants, with all the cost involved first in the edification steps and in further operation, regarding energy to reach adequate temperature and pressure. These facts show that efforts should be developed to enhance the economical attractivity of the conversion projects. The petroleum industry produces in its installations quantities of C 0 infinitely smaller than the quantities generated by the use of its products as fuels. However, it is an industrial field very intense in capital and technology, and some contributions may be expected from this industry to the development of solutions. Streams with high C 0 content can be found in refineries from the regeneration of FCC catalysts or from burning fuel oils in ovens, though pure C 0 can be obtained in good purity in an easier way by pressurization of the stream generated after absorption of acid gases in ethanolamines and elimination of sulfur present in a Claus unit; though the C 0 content is normally minor in these streams, the contaminants are basically only nitrogen and water, and C 0 can be conveniently separated. For the implementation of C 0 conversion plants in refining facilities the need for financial attraction remains, and an industrial plant to the conversion of C 0 to a given product would require a considerable price-tomarket of that sole product. In several fields, the use of C 0 in supercritical conditions has opened excellent possibilities of applications. The 5 International Symposium on Supercritical Fluids, in 2000, brought together the experiences and developments in this field in the decade of the 90's, with a rich collection of papers that showed that C 0 is by far the most studied fluid. Possible applications of C 0 in petroleum processing have directed the attention of the authors towards its use in supercritical conditions as a solvent for high value products. The conclusion of these studies, as it has been the conclusion of several researchers in the area, is that supercritical extraction with C 0 is mainly attractive as an industrial process for the extraction of high value products, like drug precursors, in small plants, due to the investment necessary. These studies have directed the authors to a different scenario for supercritical extraction and also for C 0 conversion as a way of its mitigation, porposed here. The main idea is that the energy consumed to lead C 0 to the conditions employed for the extraction of these economically attractive products could be shared with conversion plants. In supercritical extraction, the solvating capacity of C 0 is dictated mainly by the use of high pressure and temperature. After extraction, the solute is 2

2

2

2

2

2

2

2

th

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

389 normally recovered by depressurization of the extract. If suitable values for the thermodynamic parameters can be reached after depressurization, the stream of C 0 , after separation of the solute, can be directly lead to conversion plants without further energy cost. This approach has not been considered in a complete engineering project so far, but the indications available show that this is an approach able to lessen the operational cost of the units, and it can be a process solution to the attractivity of the conversion projects. In this chapter, this idea is shared with the reader as a possibility of improvement of developments in this area. After a brief discussion of the applications of extractions in petroleum industry and the advantages seen for the possible use of supercritical CO? in the generation of high value petroleum derivatives, an example is given of developments achieved by the authors with the extraction of a high value product for refineries, paraffin wax. It is taken as an evidence that works in this direction may bring profitable process solutions.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

2

Introduction The applications of supercritical fluids in processing materials have been the subject of many researches, mainly from beginning of the decade of the 90's on. In the middle of that decade some developments were already real process options, and some other ones were promising alternatives (I). By the end of the decade, though many other possibilities had been studied in addition to extraction (2, 3) only two established industrial processes could be considered: decaffeination of coffee and the so-called residue oil supercritical extraction (ROSE). The former consists of an extraction process in which supercritical C 0 is employed in substitution of liquid solvents, to avoid the residual contents present when organic solvents are employed for caffeine extraction. ROSE is an improvement of the traditional deasphalting process employed in the petroleum industry to produce valuable products from vacuum distillation residua (4). Distillation is unable to produce further fractionation of these residua for refining purposes, so extraction is employed. Both in classical deasphalting units or in ROSE plants the solvent employed (C or a mixture of C and C hydrocarbons) is not in supercritical conditions in the extraction step. In the traditional scheme the solvent is recovered by distillation. In ROSE, the extracted material is separated from the effluent solution by leading the mixture to temperature and pressure conditions above the critical point of the solvent. Under these conditions its solvating capacity is dramatically lessened and the extracted components precipitate. Deasphalting and some other extraction processes may gain more importance in petroleum industry from the end of the 90's on, because they bring possibilities for the generation of value-added products from 7

2

3

3

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

390

distillation residua. One of the main difficulties in processing many highdensity refinery feedstocks that are usual worldwide nowadays is the destination of their heavy fractions that can be very abundant after distillation and processing. Good descriptions of the need to convert distillation residua to commercial products, the installed capacity of the ROSE process and the use of approaches to the work under supercritical conditions in the extraction step can be found in literature (4). The applications of supercritical fluids in the development of new processes for petroleum refining can also be found reviewed (5,6). The applications in which potential commercial application can be seen also employ hydrocarbon solvents, as is the case in propane deasphalting, being pentane (7) or toluene (8) examples of that. The possible explanation is the obvious availability of these solvents in the petroleum and petrochemical industry and the chemical similarity with the materials to be extracted, which ensures good solvating power to these solvents. This is an important aspect because it ensures good yields of extracts, and the important product to be obtained is normally the soluble fraction. In these processes, the extracted components can be directed to applications like lubricant oils or feedstock to catalytical cracking units that produce gasoline and liquefied petroleum gas, very valuable products; the remaining fraction, consisting of very heavy hydrocarbons, can be employed as paving binders (asphalts), precursors for carbon materials or fuel oils, less valuable petroleum derivatives. That is the reason why the operation of extraction plants in petroleum industry tries to maximize extract yields. Exactly because of the similarity with the families of petroleum constituents, the use of supercritical hydrocarbons in petroleum residua (9) is not a choice for selectivity in the extracts. However, the definition of the limiting properties of the extracts required for the subsequent applications is not a function of the presence of given chemical groups; they are mainly related to some physico-chemical properties. Therefore, the requirement of selective extractions is seldom the aim of the researches in the petroleum industry. An exception is the production of paraffin waxes. These highvalue products are employed for many uses, like candles, food, and impermeabilization. They are obtained by extraction with a ketone; normally the wax-free fraction undergoes deasphalting, and that integrates the extraction plants of a refinery. However, this scheme is very dependent on the kind of petroleum employed (a very paraffinic characteristic is required), exactly because of difficulties related to selectivity in these extraction approaches. Those difficulties have motivated even the search for alternative raw materials for the production of paraffin waxes (10). Extraction with supercritical C 0 has been mentioned in literature as a good tool to achieve selectivity in extraction procedures, and since the beginning of the decade of the 90's several analytical applications based on 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

391 this selectivity aspect were already developed. Supercritical C0 had been shown to be very selective for saturates among other petroleum components (11), and by the end of that decade analytical approaches for the selective extraction of petroleum paraffins had already been proposed (12). However, studies on industrial applications were not as stablished as analytical ones. The present authors have considered that industrial applications could employ the same selectivity aspect that has brought successsful applications in analytical scale. As a first promising use of that solvent, work on the proposition of an extraction scheme for selective extraction of paraffinic materials from asphalts was developed. The interest to have asphalt as the substract of the extraction is due to the importance of the role of this product in the destination of heavy residua and also due to the fact that the extraction of paraffins matches the needs that some asphalt markets claim. In hot countries, direct action of the sun exposition may result in temperature levels in the asphaltic pavements high enough to reach the melting point of some of its components. The consequence is the fact that the rheological properties of the pavements become poorer, and deformations can be seen, mainly in curves, where a centrifuge action takes place with traffic. This problem is very well-known for asphalts, as some other ones are; identically, specification and performance problems are known to happen with other heavy petroleum products. However, it is often quite difficult to reach process solutions, because of the limitations in characterization of the components of these heavy fractions. The determination of the group of constituents to be considered responsible for the problems in the behavior of those products and, as a consequence, the proposition of process solutions for them has been a task that has brought together many efforts from petroleum chemists and process engineers. The main reason for that is the fact that heavy petroleum fractions are very difficult to be analyzed, stricto sensu, by chemical means. It is normally considered that these fractions, including asphalts, are constituted by four groups of constituents: asphaltenes (precipitated materials obtained by addition of η-heptane), paraffins, aromatics and resins (heteroatomic). These three later groups are normally obtained, based on the differences in their polarities, by preparative liquid chromatography of the whole material or of an asphaltene-free fraction. It could be said that further information has to be obtained by procedures specifically taylored to face a given problem under study. Because of these limitations, some analytical approaches aim to be specific to the properties of given groups of constituents in the heavy fractions. An example is cited for paraffins. It has been reported in literature that paraffins in asphalt are the group of its constituents that show melting transition, by studies employing fractionation and analysis of the fractions by DSC, differential scanning calorimetry (13). In these studies DSC is also employed for a quantitative estimation of paraffin contents in asphalt by the evaluation of the heat involved in the melting

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

392 transition observed in asphalt samples of known weight, with the consideration of an average melting transition value of 200 J/g for paraffinic materials. Even general analytical methods like the ones mentioned in the previous paragraph show several limitations. An example experienced by the authors can be cited. Asphaltenes are considered to be the main coke percursors in these heavy residua; however, no correlation could be found with the traditional methodology. Previous developments by the present authors had yielded quantitative and selective extraction conditions for the analytical extraction by pure supercritical C 0 (14) of paraffinic materials which had been found to co-precipitate with asphaltenes (15). A novel work was developed with paraffin-free asphaltenes, and the good performance of the methodology was validated by the improvement of correlations of asphaltene contents with carbon residue (16), a very important parameter for studies on the technologies developed for catalytical cracking of residua (17)· So, it can be seen that tayloring analytical methods are additional challenges in the development of applications of extraction fluids in petroleum industry; actually, that does not happen only for extraction monitoration, but for studies on heavy petroleum fractions in a general way, and it is a limiting aspect in the research related to those fractions. Since the attractivity of extraction processes is exactly related to the heavy fractions, that can explain the restricted implementation of extraction processes in petroleum industry so far; that has brought as a consequence very limited possibilities for C 0 utilization as a solvent. In the work presented here, where the efforts were directed to the development of conditions for selective extraction of paraffins from asphalts, a considerable fraction of the time was spent on the proposition of monitoration methods based on analytical means, and not only in specification criteria or physico-chemical properties, as it can be frequently seen in literature. The use of pure supercritical C 0 was considered a premise since the beginning of the work, since there are indications from our previous developments that the presence of co-solvents, even non-polar ones, enables extraction of non-saturates (14). So, optimization of conditions aimed basically to reach selectivity and completeness in the extractions, and though differences in the materials can be observed with variations of extraction conditions, this point will not be discussed here. In summary, extraction conditions and characteristics of both materials (soluble and insoluble ones) obtained by supercritical extraction with C 0 of asphalt will be cited. The focus of the discussion will also comprise the analytical results obtained, since that is considered a relevant contribution to the developments in this area, due to the difficulty experienced for the proposition of characterization approaches to support such works.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

393

Extraction conditions Supercritical extraction for the present study was performed in a modified Suprex Prepmaster extraction system, specially adapted for needs of this particular development. The apparatus was initially employed for analytical purposes, and it has been modified for several process studies. This has been a general tendency in the use of hardware for supercritical extraction (2). C 0 was passed through a tubular extraction cell (with ascendent C 0 flow) at the desired pressure with the aid of a reciprocant pump. That pump controlled pressure, not flow. The extraction cell was placed in an oven, for temperature control. In the original conception of the apparatus, flow was obtained by opening a 6-port valve placed in the extract exit tubing, outside the oven. A system of restriction, an aperture of reduced size, controlled the flow rate. The size of the aperture was determined by a needle. Since flow variations happen during extraction process, this restrictor was controlled by an engine. So, it was built outside the body of the extractor. The collection of extracted material was done onto adsorbents, conditioned in a small metallic cylinder. Refrigeration gas could be admitted outside this cylinder wall to improve retentioa Paraffin depositions was a difficulty found in the operation of that original construction. So, some modifications were necessary to avoid that problem. AU of the extract exit part of the apparatus was removed. Two slightly open needle valve were placed and they allowed the system to be under high pressure at the same time in which the extract was being collected; the use of two valves is explained by the fact that an only valve is unable to control the desired flow (measured by the displacement of the pump strokes) in the conditions at high pressure. After leaving the extraction cell, the C 0 stream was lead to the bottom of a column of glass beads; there C 0 was at room pressure again, going to the gas state, and the extracted material precipitated, being mechanically retained by the glass beads. Analytical approaches were the next step considered in the development. The first aspect to be considered before characterization of the extracts, that is, the evaluation of extraction yields, was monitored by gravimetric means. Since the mass of asphalt placed in the extraction cell was always known, the extracted fraction collected in the glass beads and the remaining insoluble material were both weighed after each extraction, and the mass balance was checked. Extraction temperature and C 0 flow employed were obtained from previous experience (14). Even high carbon-numbered standard paraffins are soluble enough at temperatures inferior to 120°C, so that was the minimum temperature considered. It will be seen that large variations in temperature were not possible, because of asphalt nature. The previous experience also showed that the C 0 flow is the variable that dictates the time for extraction, since diffusion is not the aspect that controls the final 2

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

394

amount in solution; the thermodynamic equilibrium is easily reached. So, only pressure was considered a variable to be explored in this study.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

Analytical Monitoration of Paraffin Extraction of Asphalts As it was mentioned before, the analysis of asphalts is a very difficult task, since the commonly employed analytical techniques in petroleum industry, like gas chromatography, show many limitations for application to the characterization of any of the heaviest fractions of petroleum (18). Characterization schemes of both the extracted fraction and the insoluble material were proposed mainly with use of thermal analysis. DSC was employed in a diverse manner of the previously cited application, in which an average heat of fusion was employed to a quantification of paraffin content (13). In this work, two kinds of use were reserved for DSC graphs. The first one aimed to evaluate the result of continuously decreasing amounts of paraffins remaining in the asphaltic material by the increase of the extraction time. DSC curves, therefore, only brought a tendency of the softening point of the novel asphalts produced by paraffin removal, which is an interesting indication related to the properties of the extracted product. The second (and less frequent in this work) one was the monitoration of the few experiments in which quantitative removal of paraffins was desired, that aimed to yield an idea of the behavior of paraffin-freed asphalts. So, extraction was performed in these cases till a baseline was obtained in the DSC graphs. Actually, in the beginning of the work, the heat flow used for melting remaining paraffins in the sample after extraction was compared to the value observed before extraction to be used as a quantification of remaining paraffins in asphalt. The results showed at once that neither this value nor the value of 200 J/g (13) employed for asphalt analysis can be trusted for quantification purposes, since the amounts of the fractions of paraffinic materials yielded by gravimetric evaluation of the preparative extraction never matched the one calculated with that average transition heat. This is probably due to the fact that the heat involved in that transitions is a function of the distribution of methyl and methylene groups in the aliphatic chains of the paraffins (19). That distribution probably changes with the severity of the extraction, since the number of paraffinic components in asphalt is quite large, and different solubility behaviors can be expected. That makes it impossible to employ calorimetric measurements for these quantitative estimations, unless only semiquantitative results are expected. So, DSC measurements were only employed for the monitoration of the new transition onset temperatures obtained for the extracted asphalts. In some few cases where quantitative preparative extractions were desired,

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

395 the curves were also employed to the evaluation of the completeness of the extraction, which could be observed by the disappearance of the transition peak. Another thermal analysis technique, thermogravimetry (TG), was used for rapid evaluation of the presence of co-extracted materials in the paraffinic fractions obtained, since selective extracts were sought. The application was based on the knowledge that paraffinic materials would not yield coke when heated under inert atmosphere (20). So, a ready way to detect how selective a given extraction condition had been was to perform TG analysis up to 70Q°C under nitrogen. The presence of residue (coke) at that temperature indicated the presence of non-paraffinic material in the extract. Coke formed by the pyrolysis of those components could be burned by the introduction of air in the system at 700°C, and additional mass loss was observed. Additionally, nuclear magnetic resonance (NMR) and elemental analysis (CHN) were also employed. These are typical analysis for heavy petroleum products, so they will not be longly discussed here. Briefly, it can be cited that *H NMR spectra show baseline elevation in the chemical shift range of 9 to 6 ppm when aromatic hydrogens are present in petroleum or its fractions (21). The atomic ratio (C/H) of long-chained paraffins is 0.5, and the presence of meaningful amounts of aromatics may be detected with any deviation to a higher value. It should be also mentioned that some other analytical approaches were tried in the beginning of the developments. The first technique considered to monitor the extraction performance was gas chromatography. However, elution could not be confirmed as quantitative, and resolution enough for such a heavy and complex system could not be achieved. Normal phase liquid chromatography was also considered, but separation of asphaltenes before injection would be necessary, to avoid irreversible adsorption onto stationary phase. Since that could bring doubt to the results expected, because of the possibilities of co-precipitation of the heavier paraffins (15), and the whole procedure would become very tedious and time-consuming, further investigations on this technique were not performed.

Experimental Results for Extraction The first critérium employed in the monitoration of the extractions was the heat flow employed for the melting transition of the remaining paraffins in the asphalt. This product was called modified asphalt. Figure 1 shows typical DSC curves of the original and modified asphalt; in this example, extraction was performed at 360 atm. The properties of the departing asphalt are cited in table 1.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

396 Table 1. Main properties of the asphat studied Method Parameter ASTMD-2171 Viscosity at 60°C, Ρ ASTM D-5 Penetration, 100g, 5 s, 25°C, tenths of mm ASTMD-92 Flash point

Found 2800 55 238

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

Before extraction

After extraction

—koTî^— ι

• • •

.

1 50

0

I

I

100

Q

C

Figure 1. DSC graphs ofasphaltic material before and after supercritical extration with CQ at 360 aim and 150*0 (CO2 flow was 16mL/min) 2

From Figure 1, it can be seen that it was possible to achieve practically paraffin-free asphaltic material. Quantitative extraction is not adequate for paving purposes. Very high softening points make the application on the pavement quite difficult. It can be also seen in Figure 1 that there is an elevation in the baseline after the paraffins melt. This means that the sample can be undergoing chemical modification. So, the maximum temperature to be reached could not be superior to about 120°C, and that was the adopted value for this parameter. The data in Figure 1 bring some additional important aspects. Only the heaviest paraffinic components are not extracted, but their melting temperature range onset (related with softening) is about 100°C. This does not bring any mechanical problem for the utilization of the modified product obtained as asphalt binders; some problems would probably be verified in the use of the original material in many hot countries, since its onset is below 40°C, and that temperature level can be easily reached in summer in these places. Though DSC measurements are very informative about characteristics of insoluble fraction, they could not inform about the selectivity of extraction. Characterization of nature of paraffinic fraction obtained was done by *H NMR and elemental analysis for C, Η and N. Selectivity for saturates was clearly evaluated from results obtained by these two techniques. Nitrogen was absent in extracts. Atomic ratio C/H was almost exactly 0.5, while weight percent contents of C and Η summed In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

397 practically 100%. NMR spectra could not show presence of aromatic hydrogens in extract (Figure 2), while for the whole asphaltic material (Figure 3) the presence of aromaticity could be evaluated in 13% of the total hydrogen content. The single peaks at about 7.2 ppm in Figures 2 and 3 are due to the presence of CHC1 in the analysis solvent, CDC1 .

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

3

3

_ I

l

9

I

8

I

7

6

ι

5

ι

4

ι

3

J ι

2

ι

1 PPm

1

Figure 2. H NMR spectrum of the extract obtained with supercritical CO2

τ—ι—ι—ι—ι—r—1—1—1—1—1—1—

11 10 9

8

7

6

5

4

3

2

1 -0 PPm

Figure 3. H NMR spectrum of the whole asphaltic material before supercritical extraction 1

l

The R NMR spectrum in Figure 3 also shows that are considerable amounts of aliphatic branches bonded to aromatic nuclei in the asphalts; that can be evaluated by the meaningful envelope from 4 to 2 ppm (21). It was considered in the beginning of the work that these branched components could be a challenge for selectivity, because of the In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

398

possibility of a hybrid behavior of the ones branched with long aliphatic chains. In Figure 2 it can be seen that C 0 could be very selective. Additional characterization of extracted material was achieved with thermogravimetry. Figures 4 and 5 show the difference between thermal behaviors of the asphaltic material and the extract.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

2

0

200

400

600

800

1000

a

C

Figure 4. TG graph of the whole asphaltic material

0

200

400

600

800

1000

Figure 5. TG graph of paraffinic extract

Thermal behaviors show that structures in the original asphaltic material yield large amounts of coke (approximately 50%), while the extract does not; that is, after introduction of air. there is no considerable mass loss. This was taken as additional evidence of the selectivity of extraction for saturates.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

399

Extraction yields

Extraction yields vary with the conditions in which the extraction takes place. At 120°C, no extraction is achieved when the C 0 pressure is 50 atm; a fraction of light paraffin is obtained when the pressure is about 200 atm; quantitative extraction is achieved when the pressure is 400 atm. Figure 6 shows the tendency observed with a C 0 flow of 16 mL/min, after a 3 hour extraction time. 2

14η weighit % extracted

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

2

after a 3 hour extraction time.

1211864-

y= 6,1449Ln(x) - 23,742 R2= 0,9879

S

f

2n C)

100

200

300

400

500

Extraction pressure (atm) Figure 6. Extraction yield tendency with solvent variation This means that the extraction conditions can be planned to reach different materials, depending on the need to produce micro or macrocrystalline paraffins; the characteristics of these fractions are mainly a function of the molecular weight range (19), and since that range also defines the solubility, the extraction is an adequate way to obtain separately those products. This also means that if the C 0 stream pressure after extraction is reduced to 50 atm, the paraffin wax precipitates, and C 0 , after generation of two value-added products, can be directly lead to conversion plants, without further energy costs on feedstock preparation to those plants. The pressure level mentioned here (50 atm) is adequate to most of the conversion developments seen in literature, many of them cited in this publication. 2

2

Conclusions Supercritical extraction with pure C 0 may yield valuable products, as it was seen here in the examples of paraffin wax and paraffinfree asphaltic material. This chapter reported the necessary steps employed for a successful development of conditions for selective and complete 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

400 extraction of paraffins from asphalts. Since heavy petroleumfractionsare the possible field of application of extractions, it was seen that characterization schemes play a vey important role in the development of extraction conditions, because of difficulties in chemical characterization of the components of thesefractions.This is expected to be a contribution to novel developments and to a more popular use of C 0 in petroleum industry. The optimization revealed conditions where the solubility of the material is null. If the extract is lead to those conditions, the extracted material precipitates and the C 0 stream generated can be obtained in adequate temperature and pressure to be used in some further applications without additional energy consumption; it is proposed here that this stream can be a ready feedstock for conversion plants. The applicability of supercritical C 0 opens large possibilities for those coupled processes, since investments and operational costs can be shared by two industrial units. The conversion to methanol, for example (22), can be performed with C 0 at 50 atm. At that pressure, all of the extract content would be precipitated after depressurization, so that the resulting C 0 stream could be lead directly to the conversion plant. About the economics of the process, it is very difficult to state about that without a whole engineering project. However, it can be cited that the values mentioned in literature (22) indicate that the cost of C 0 conversion to methanol can be calculated in about US$50,00 per ton C 0 . That is a meaningful value, so that a previous use of C 0 for the obtention of a product in the range of at least US$750,00/ton, as paraffin wax is, is certainly an attractive option to be considered. 2

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

2

2

2

2

2

2

2

References 1. 2.

Lynch, T. J. Chem. Tech. Biotech. 1996, 65 (3), 293. Perrut, M . Proceedings of The 5 International Symposium on Supercritical Fluids, Atlanta, Georgia, April 8 -12 , 2000, Plenary Lecture 2. Chordia, L.; Robey, R. Proceedings of The 5 International Symposium on Supercritical Fluids, Atlanta, Georgia, April 8 -12 , 2000, Plenary Lecture 3. Kim, M . S.; Yang, K. S.; Hwang, J. S. Pet. Sci. Tech. 1997, 15, (9/10), 921. Rudzinski, W. E.; Aminabhavi, T. M. Energy & Fuels 2000, 14, (2), 464. Dadashev, M . N.; Stepanov, G. V. Chem. Tech Fuels & Oils 2000, 36, (1), 8. Stegeman, J. R. et al. Fuel Sci. Tech. Int., 1992, 10, (4-6), 767. Zhuang, Mark S.; Thies, Mark C. Energy & Fuels 2000, 14(1), 70. Deo, M. D.; Hanson, F. H. Fuel 1994, 73 (9), 1493. th

th

3.

th

th

th

4. 5. 6. 7. 8. 9.

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

th

401 10. Badyshtova, K.M.; Shabalina, T.N. ; Kitova, M.V. Chem. Tech Fuels & Oils 1996, 32(3), 119. 11. Singh, I. et al. Fuel Sci. Tech. Int. 1992, 10(2), 267. 12. Oschmann, H.J.; Prahl, U.; Severin, D. Pet. Sci. Tech. 1998, 16,(1/2), 133, 13. Burlé, Β. et al. Symposium on Chemistry and Characterization of Asphalts Proceedings, 1990, 330. 14. Teixeira, M . A. G.; Oliveira, L. B., Proceedinds of the 3 Brazilian Meeting on Supercritical Extraction of Natural Products, Rio de Janeiro, 1999, Paper#60. 15. Teixeira, M . A. G.; Gonçalves, M . L . Α., Petroleum Science and Technology 2000, 18(3-4), 273. 16. Teixeira, M . A. G.; Proceedings of The Rio Oil and Gas Expo and Conference, 2000, Paper#250. 17. Baptista, C. Μ Α.; Bonfadini, P. R.; Gilbert, W. R.; Teixeira, M . A. G. Proceedings of the National Petrochemical & Refiners Association Annual Meeting , San Antonio, 2000 , paper #AM -00-23. 18. Teixeira, M . A. G.; Faria, F. R. D.; Albuquerque, F. C. , Proceedings of the Second International Symposium on Colloid Chemistry in Oil Production, 1997, Paper#44. 19. Gupta, A. K.; Brouwer, L.; Severin, D. Pet. Sci. Tech. 1998, 16(1-2), 59. 20. Teixeira, M . A. G.; Gonçalves, M . L. Α., Petroleum Science and Technology, 1999, 17(1&2), 1. 21. 12. Hassam, M. H. et al. Fuel 1983, 62. 22. Steinberg, M. American Chemical Society, Division of Petroleum Chemistry, Inc. Preprints, International Symposium on CO Conversion and Utilization in Refinery and Chemical Processing, 219 American Chemical Society Meeting, San Francisco, 2000, 45, (1), 74.

Downloaded by COLUMBIA UNIV on September 16, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ch025

rd

2

th

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix001

Author Index Nakayama, M . , 344 Ogura, K., 344 Ohtsuka, Yasuo, 85 Omata, K . , 153 Pan, Wei, 258,316 Scaroni, Alan W., 166, 258 Sekine, Yasushi, 303 Shamsi, Abolghasem, 182 Song, Chunshan, 2, 166, 258, 289, 316 Srimat, Srinivas T., 258, 289 Steinberg, M . , 31 Subramaniam, Bala, 364 Sun, Lu, 258 Suzuki, Toshimitsu, 205 Tan, Chung-Sung, 102 Teixeira, Marco Antonio G., 387 Tomishige, Keiichi, 71, 303 Uphade, B . S., 224 Ushizaki, K . , 153 Wang, Ye, 85 Wei, Jun-Mei, 197 X u , Bo-Qing, 197 Yamada, M . , 153 Yamashita, Hiromi, 330 Yano, H , 344 Yoo, JinS., 112 Yoshida, T., 241,275 Yoshinaga, Yusuke, 303 Zhang, Z.-G.,241,275 Zhu, Qi-Ming, 197

Akamatsu, Noriyasu, 205 Anpo, Masakazu, 330 Aresta, Michèle, 54 Armor, John N . , 258 Asadullah, Mohammad, 303 Babcock, R. E., 224 Busch, Daryle H., 364 Chang, Char-Fu, 102 Cheng, Zhen-Xing, 197 Choudhary, V . R., 224 Dibenedetto, Angela, 54 Effendi, Α., 275 Fujimoto, Kaoru, 71, 303 Furusawa, Yutaka, 71 Gaffhey, Anne M . , ix Gonçalves, Maria Luisa Α., 387 Haraguchi, K., 241 Hideshima, Shigeo, 205 Hsu, Tsung-Ju, 102 Ikeda, Yoshiki, 71 Ikenaga, Na-oki, 205 Ikeue, Keita, 330 Inui, Tomoyuki, 130 Ishiguro, G., 153 L i , Jin-Lu, 197 Mahajan, D., 166 Mamman, A . S., 224 Manzer, Leo E., 39 Matsui, Na-oko, 205 Matsuo, Yuichi, 303 Murata, Satoru, 258 Nakagawa, Kiyoharu,205

403 In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

Subject Index Acrolein, propylene oxidation, 42 Acrylic acid oxidation of propane, 44t propylene oxidation, 42 Acrylonitrile, ammoxidation of propane, 44/ Adsorption, separation of gas mixtures, 10 Aerobic oxidations vs. anaerobic, 40-43 See also Catalytic oxidations Alcohol, allylic and homoallylic, oxidation in dense phase C 0 , 374 Alcohol synthesis alcohol distribution for different catalysts, 143, 145/ catalyst design, 140, 143, 147 comparison of temperatureprogrammed-reduction (TPR) profiles for various catalysts, 143,144/ ethanol, 140 ethanol synthesis from C 0 on various catalysts, 142/ Schulz-Flory plots for different catalysts, 143, 146/ See also Catalytic conversion Aldehydes. See Gas-phase 0 oxidation of alkylaromatics Alkenes, catalytic oxidation in scC0 , 374, 375/ Alkylaromatics. See Gas-phase 0 oxidation of alkylaromatics Alkylation reactions, dense phase C 0 , 368-369

Ammoxidation, propane to acrylonitrile, 44/ Anaerobic oxidations vs. aerobic, 40-43 See also Catalytic oxidations Aromatics. See Gas-phase 0 oxidation of alkylaromatics Asphalt analytical of paraffin extraction, 394-395 asphaltenes, 392 C NMR spectrum of material before extraction, 397/ constituents, 391 differential scanning calorimetry (DSC) before and after supercritical extraction, 396/ extraction results, 395-398 extraction yields, 399 H NMR spectrum of extract, 397/ main properties, 396/ paraffins, 391-392 thermogravimetry of asphaltic material, 398/ thermogravimetry of paraffinic extract, 398/ Atmospheric pressure. See Pressure effects 2

13

2

l

2

Β

2

Benzene, dinitrogen oxide oxidation, 47/ Binary catalysts. See Methane conversion Binary oxide, Ti/Si, 341 Biotechnology

2

2

2

404

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

405 coupling chemistry with, 66-68 synthesis of 4-OH-benzoic acid, 68 Boudouard reaction carbon deposition, 208, 210 pressure drop across catalyst bed, 237 Brookhaven National Laboratory, integrating C 0 sequestration with conversion in geologic formations, 178-179 Building blocks. See Synthetic chemistry Burial, sequestration option, 167 Butadiene, oxidative dehydrogenation of butane, 42 Butane, oxidative dehydrogenation to butadiene, 42 Butane oxidation maleic anhydride, 42-43 selectivity, 40/ See also Catalytic oxidations

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

C Ca-based binary catalysts. See Methane conversion Capture, C 0 , 9-10 Carbamates, synthesis, 65 Carbide catalysts methane dry reforming, 185/ preparation, 185-186 See also Methane dry reforming Carbonates synthesis, 62, 64-65 synthesis routes, 72-73 See also Dimethyl carbonate (DMC) Carbon dioxide, C 0 area for C 0 control, 8/ barriers for conversion and utilization, 20-21 capture and sequestration, 9-10 2

2

2

challenges and strategies for conversion and utilization, 16, 20-24 chemical processes for conversion and utilization, 19/ control of emissions, 6, 9 conversion and utilization, 10-16 current status and potential market for utilization, 14, 16 estimates for utilization, 18/ estimates for worldwide capacity of options for sequestration, 18/ extraction by Carnol process, 32 Gibbsfreeenergy of formation, 13/ green chemistry, 112 greenhouse gases (GHG), 3, 8/ key issues for control of greenhouse gas, 8/ mild oxidant with chromiumbased catalysts, 49, 50/ occurrence, 2 properties, 12t, 365 reaction with hydrogen, 33 sources of emissions, 3, 6 stationary, mobile and natural sources, 4/ strategy for research on utilization, 21-24 technical options for emission control, 6 thermodynamics of conversion, 11, 14 U.S. emissions from different sectors, It U.S. emissions from electricitygenerating units, It uses, 11 U.S. production of liquid fuels in 1999, 17/ U.S. production of synthetic plastics and related chemicals (1999), 15/

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

406 utilization with and without chemical conversion processing, 2 2 / world emissions from consumption of fossil fuels (1980-1997), 5/ See also Catalytic reduction of C 0 ; Dense phase carbon dioxide; Dimethyl carbonate (DMC); Simultaneous C 0 and steam reforming of methane; Synthetic chemistry Carbon dioxide mitigation. See Carnol process Carbon supported cobalt catalysts. See Methane decomposition Carboxylation mechanism, 58 organic substrates, 60-62 Carnol process carbon dioxide extraction, 32 C 0 emission evaluation, 34, 36/, 38 C 0 from coal burning, 32-33 economics, 34 entire Carnol system, 37/ hydrogen production, 32 methanol and higher oxygenates as liquid automotive fuel, 3 3 34 reacting hydrogen with C 0 , 33 thermodynamics, 35/ Catalyst, N i / S i 0 - M g O . See Methane reforming with C 0 over N i / S i 0 - M g O Catalysts, Ca-based binary. See Methane conversion Catalysts, Ni/Na-Y and N i / A l 0 . See Methane reforming with C 0 over Ni/Na-Y and N i / A l 0 Catalysts, supported Cu and Mn. See Methanol synthesis from C 0 containing syngas Catalytic conversions 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

2

2

2

2

2

2

3

2

2

2

3

activity of C 0 methanation for N i - L a 0 - R h catalyst, 137/ aimed reactions producing only syngas, 133 alcohol distribution for different catalysts, 145/ alcohol synthesis from C 0 - H mixture, 140, 143, 147 applying N i - L a 0 - R h catalyst to higher space velocity of reaction gas, 137-138 catalysts for alcohol synthesis, 143,147 C0 -methanation and global warming, 136 C0 -methanation for Ni-based three-component composite catalyst, 136-137 comparison of temperatureprogrammed reduction (TPR) profiles, 143, 144/ CO- and C0 -rich syngas, 147, 149 C u - Z n - C r - A l mixed oxides catalyst by uniform gelation method, 138 effect of catalytic oxidation of ethane and propane on C 0 reforming of methane, 135/ effect of G a 0 addition to four component catalyst, 140 endothermic heat of reaction, 132 ethane or propane addition in C0 -reforming of methane, 135 hydrocarbon synthesis from syngas, 148/ methanol synthesis activity and combination of L a 0 and Pd or Ag, 138, 140 methanol synthesis by C 0 hydrogénation, 138-140 methanol synthesis catalyst design considerations, 173, 175 methanol synthesis for C 0 - and 2

2

3

2

2

2

3

2

2

2

2

2

3

2

2

3

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

2

407 CO-rich syngas on fourcomponent catalyst, 141/ methanol synthesis from C 0 on various catalysts, 139/ Ni-based three-component catalyst, 132-133 Pd-Ga-modified catalyst and COrich syngas conversion, 140 rapid C0 -methanation, 135-138 rapid C0 -reforming of methane, 131-135 reviews, 132 reviews of C0 -methanation, 136 Rh addition to N i - C e 0 - P t catalyst, 133, 134/ Schulz-Flory plots for alcohol synthesis, 146/ selective synthesis of light olefins and gasoline from C 0 - H mixture, 147-149 superior to other C 0 conversion methods, 131 synergistic activity of composite catalyst, 133 synergistic effect of composite catalysts and combined reactions, 132-133 synthesis of ethanol from C 0 on various catalysts, 142/ Catalytic oxidations alternative oxidants, 46-49 ammoxidation of propane to acrylonitrile, 44/ anaerobic vs. aerobic, 40-43 carbon dioxide as mild oxidant with chromium-based catalysts, 49 chemistry under unusual conditions, 45-46 cost of oxygen sources, 47/ coupling methane to ethylene, 41-42 incentive for direct oxidation of paraffins, 4 3 / 2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

2

3

2

2

2

2

Lummus process for ammoxidation of o-xylene to ophalonitrile, 41 Mars-van Krevelen oxidation of butane to maleic anhydride, 4 1 / N 0 as selective oxidant for aromatics, 46^17 new methyl methacrylate process, 45 new process chemistry, 45-46 nylon intermediates manufacturing, 47 oxidation of butane to maleic anhydride, 42-43 oxidation of propane to acrylic acid, 44/ oxidation of propylene to acrolein and acrylic acid, 42 oxidative dehydrogenation of butane to butadiene, 42 oxidative dehydrogenation of ethane to ethylene, 46/ oxidative dehydrogenation of ethane with C 0 , 4 9 / 50/ oxidative dimerization of isobutylene to 2,5dimethylhexadiene (DMH), 42 oxidative dimerization of toluene to stilbene, 42 paraffin oxidations, 43-44 preparation of propylene oxide from hydrogen/oxygen mixtures, 4 9 / proposed mechanism for N 0 oxidation of benzene, 4 7 / schematic of two-stage oxidation of o-xylene, 4 1 / selectivity, 39-^0 selectivity of chromium-based catalysts, 49/ titanosilicate oxidations, 4 8 / titanosilicates, 47-48 typical selectivity, 40/ Catalytic reduction of C 0 2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

408 basis for proposing geologic formations as slurry reactors, 176-178 batch unit, 168 carbon management, 167 catalyst evaluation studies, 168169 catalytic hydrogénation screening runs, 175/ C 0 solubility data in various solvents, 171/ C 0 solubility studies, 169, 171173 effort at Brookhaven National Laboratory (BNL) and Pennsylvania State University (PSU), 178-179 elements of catalytic system converting sequestered C 0 to liquid fuels, 177 endothermic transformation to liquid fuels, 167 equilibrium pressure-gas mol fraction isotherm for C 0 solubility in PEG-400 at room temperature, 172/ evidence for catalytic effects of certain metals, 177-178 experimental, 168-169 general scheme for aquifer/ocean sequestration of C0 , 170/ geologic formations, ultimate slurry reactor, 175-179 integrated approach to catalytic C 0 hydrogénation, 169-175 liquid phase catalytic hydrogénation ofC0 /CO, 174/ methanol synthesis catalyst design considerations, 173, 175 oil and gas reservoirs, 176 proposed integrated system for production of ultra clean fuels from C0 , 170/ sequestration options, 167 solubility studies, 168 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

2

2

2

2

2

temperature and pressure gradients of earth, 176-177 See also Liquid fuels; Photocatalytic reduction of C 0 Chemical conversion processing, utilization of C0 , 21-24 Chemical processes, C 0 conversion and utilization, 19/ Chemical vapor deposition (CVD) Fe/Mo/DBH catalysts Ag-doping, 121 silica modification, 120-121 See also Gas-phase 0 oxidation of alkylaromatics Chromium-based catalysts, carbon dioxide as mild oxidant with, 49, 50/ Coal, Carnol process, 32-33 Cobalt catalysts, carbon supported apparatus, 243 catalytic activity, 250, 252 catalytic activity method, 244 characterization, 244-250 characterization methods, 243244 Co dispersion, 245, 247, 250 Co state and particle size, 245, 246/ effect of reduction conditions on relative CO dispersion, 247, 250,251/ H production or carbon supported catalyst at 300 or 350°C, 253/ hydrogénation experiments, 252 molar ratio of produced H to decomposed C H vs. reaction time, 254/ physical characteristics, 244-245, 246/ preparation, 242-243 profiles of concentrations of produced H and unreacted CH , 251/ temperature programmed

2

2

2

2

2

2

4

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

409 reduction (TPR) profiles for methane decomposition, 248/ TPR patterns of hydrogen obtained, 249/ typical profile of CH , C H , and C H production vs. hydrogénation time, 254/ See also Methane decomposition Cobalt-nickel oxide catalysts catalyst characterization, 227 catalyst characterization methods, 226 catalytic reactions for C 0 reforming and simultaneous steam and C 0 reforming of methane, 226 C 0 reforming of methane, 229 curves for temperature programmed reduction (TPR) of catalyst with different Co/Ni ratios, 228/ effect of cobalt addition to catalyst, 233, 237 experimental, 225-226 influence of Co/Ni ratio on catalyst performance in simultaneous C 0 and steam reforming, 233, 234/ influence of Co/Ni ratio on conversion, H selectivity, and H /CO product ratio, 231/ influence of Co/Ni ratio on pressure drop across catalyst bed, 230/ influence of temperature, space velocity, and C H / C 0 feed ratio, 232/ optimum CO/Ni ratio, 237 performance of catalyst with optimum Co/Ni ratio on simultaneous C 0 and steam reforming, 235/ 236/ practical importance of simultaneous C 0 and steam reforming process, 237-238 4

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

3

2

4

8

2

2

2

2

2

2

4

2

2

2

preparation of supported Co Nii_ O, 225-226 pressure drop across reactor, 237 simultaneous C 0 and steam reforming of methane, 233 surface area and degree of reduction (in TPR) of supported catalysts, 227/ CO disproportionation carbon formation during C 0 reforming of CH , 269-270 equation, 276 Coking applications of in situ decoking concept, 368 carbonaceous species along fixed-bed, 285, 287 characterization of nature, 282, 285 deposited, 282, 285, 287 determination method, 278 equations, 300 evidence of rapid catalyst deactivation in fixed-bed, 282, 283/ graphitic carbon deposits, 354, 357 origin, 269-270 resistance of noble metals, 290 See also Methane reforming with C 0 over Ni/Si0 -MgO Computational analysis of energy of methane reforming comparison of effects of C 0 / C H and 0 / C H ratio on reaction heat at different pressures, 326, 327/ comparison of effects of C 0 / C H and 0 / C H ratios on reaction heat at different temperatures, 321, 324/ 325/ computational method, 317-318 effect of C 0 / C H and 0 / C H ratio in feedstock on reaction heat under 1 atm, 322/ 323/ x

x

2

2

4

2

2

2

4

2

4

2

4

2

4

2

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

4

410

effect of reaction temperature and pressure on reaction heat of C 0 + C H = 2CO + 2H , 318320 effects of 0 /CH and C0 /CH ratios on reaction heat, 320-321 endothermic reactions, 320 exothermic reactions, 320 trend of changes in reaction heat with temperatures and pressures, 319/ Conversion barriers for C 0 , 20-21 challenges and strategies for C0 , 16, 20-24 chemical processes, 19/ C 0 , 10-11 energy requirements, 20 thermodynamics of C0 , 11, 14 See also Catalytic conversions; Methane conversion Copolymerization of C 0 with epoxides C N M R spectrum, 107/ C 0 with propylene oxide (PO), 103 component ratios in rare-earthmetal coordination catalyst system, 109 copolymerization scheme, 105 effect of monomer ratio on yield and molecular weight, 109/ effect of pressure, 108/ effect of solvent, 104, 105/ effect of temperature, 108/ epoxides PO and cyclohexene oxide (CHO), 103-104 experimental, 103-104 H NMR spectrum, 107/ infrared spectrum, 106/ results for different yttrium coordination catalyst systems, 105, 106/ screening of solvents, 104, 105/ 2

4

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

4

2

4

2

2

2

2

2

13

2

ternary rare-earth-metal coordination system, 103 yttrium-metal coordination catalyst, 103 Copper catalysts Cu-loaded T i 0 powder, 333-334 See also Methanol synthesis from C 0 containing syngas Copper plate electrodes. See Electrochemical reduction of C0 CO reduction, pressure drop across catalyst bed, 237 Costs barrier for C 0 conversion and utilization, 20 oxygen sources, 47/ See also Economics Critical phase behavior, fluid mixtures, 367 Cycloalkanes, reactivity towards C0 , 63 Cyclohexene homogeneous oxidation in scC0 , 373-374 hydrogénation on supported Pd catalysts using ,scC0 , 370371,372/ oxidation with fluorinated iron porphyrin catalyst, 379, 380/ oxidation with non-fluorinated iron porphyrin catalyst, 379 Cyclohexene oxide. See Copolymerization of C 0 with epoxides 2

2

2

2

2

2

2

2

l

Deasphalting process petroleum industry, 389-390 propane, 390 residue oil supercritical extraction (ROSE), 389

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

411 Decaffeination of coffee, supercritical C 0 , 389 Dense phase carbon dioxide advantages over conventional solvents in homogenous catalytic oxidation, 382 alkylation reactions, 368-369, 372/ ^ alleviation of internal porediffusion resistances, 368-369 applications, 365 applications of in situ decoking concept, 368 catalytic oxidation of alkenes in

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

scC0 , 374, 375/ 2

common conditions for homogeneous catalytic 0 oxidation of di-/er/-butylphenol (DTBP), 378-379, 380/ contrasting temperature dependence of conversion to products vs. selectivity, 377 critical phase behavior of fluid mixtures, 367 eliminating inter-phase mass transfer resistances, 369-373 expanding ionic liquids with s c C 0 , 383 expanding solvents by adding C 0 , 378 heterogeneous catalytic reactions, 367-373 homogeneous oxidation catalysis of C H R e 0 and olefin epoxidation in 5 c C 0 , 374, 376 homogeneous oxidations in s c C 0 , 373-377 hydrogénation of cyclohexene on supported Pd catalysts, 370371,372/ hydrogénation reactions, 369-370 industrial applications in heterogeneous fluid-solid catalysis, 383 2

2

2

3

3

2

2

Jacobson's catalyst for epoxidation of olefins, 376 limitations of scC0 , 377-378 miscibility of oxygen with scC0 , 376 oxidation of cyclohexene with fluorinated iron porphyrin catalyst, 379, 380/ oxidation of cyclohexene with non-fluorinated iron porphyrin catalyst, 379 oxidations, hydroformylation, and esterification, 371, 373 oxidations in C0 -expanded solvent media, 377-381 oxidations of di-/er/-butylphenol, 376-377 pressure effects, 366-367 pressure-tunability of density and transport properties of nearcritical fluids, 381-382 selective oxidation of activated alkenes, allylic and homoallylic alcohols, 374 thermodynamic and kinetic aspects, 366-367 Dialkoxydibutyltin, dimethyl carbonate (DMC) formation, 73 Dimethyl carbonate (DMC) alcohols with urea, 72-73 dependence of D M C amount and surface area of H P 0 / Z r 0 calcined at 673 Κ on H P 0 loading, 75, 7 8 / dependence of D M C amount and surface area on calcination temperature of H P 0 / Z r 0 , 79, 80/ dependence of D M C amount on reaction temperature over Z r 0 and H P 0 / Z r 0 , 75, 7 8 / effect of H P 0 / Z r 0 preparation method on D M C formation, 81, 83/ 2

2

2

3

4

2

3

3

4

4

2

2

3

4

3

2

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

412 effect of modified Z r 0 catalysts calcined at 673 Κ and 925 Κ on D M C formation, 75, 77/ ethylene oxide route, 73 experimental, 74 large-scale production methods, 72-73 L R S spectra of fresh H P 0 / Z r 0 by different preparation methods, 81, 82/ methanol and C 0 with dialkoxybutyltin, 73 methanol and C 0 with zirconia catalysts, 73-74 methanol and phosgene in sodium hydroxide, 72 oxidative carbonylation of methanol, 72 oxidative carbonylation using palladium and methyl nitrite, 72 preparing H P 0 / Z r 0 by different method, 79 results of methanol and carbon dioxide over heterogeneous catalysts, 75, 76/ supercritical C 0 and trimethyl orthoacetate, 73 X R D patterns of fresh H P 0 / Z r 0 by different preparation methods, 79, 8 1 / X R D patterns of fresh H P 0 / Z r 0 calcined at various temperatures, 79, 80/ X R D patterns of fresh H P 0 / Z r 0 with various H P 0 loadings, 75, 79/ 2,5-Dimethylhexadiene, oxidative dimerization of isobutylene, 42 Dinitrogen oxide ( N 0 ) nylon intermediates manufacturing, 47 proposed mechanism for N 0 oxidation of benzene, 4 7 / selective oxidant for aromatics, 46-47 2

3

4

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

4

2

2

3

4

2

3

4

2

3

4

2

3

4

2

2

2

3

See also Catalytic oxidations Di-/er/-butylphenol (DTBP) homogeneous catalytic ep­ oxidation, 378-379, 380/ oxidations in s e C 0 , 376-377

2

2

Economics Carnol process, 34 See also Costs Electrochemical reduction of C 0 black film on surface of electrodes, 354, 357 C H and C H generation by graphite reduction, 357 chemical fixation of C 0 , 345 copper/gas-diffusion and copper plate electrodes, 348, 354, 357 Cu electrode in hydrogen carbonate solution, 345 Cu/gas diffusion and Cu plate electrodes vs. potential, 349/ electrolysis cell with gasdiffusion electrode, 347/ equations for formation of C H and C H , 354 experimental, 346-348 Faradaic efficiencies with Cu/gas-diffusion electrode, 351/ Faradaic efficiencies with Cu plate electrode, 350/ graphite formation with copper oxide involvement, 357 major products in two electrolysis systems vs. potential, 352/ 353/ modified Pt/gas-diffusion electrode, 357-359 products of modified Pt/gasdiffusion electrode, 358/ reduction scheme on polymermodified electrode, 359, 360/ total concentration of products vs. 2

4

2

4

2

2

4

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

4

413 electric charge on Cu/gasdiffusion electrode, 356/ total concentration of products vs. electric charge on plate electrode, 355/ Electrodes, gas-diffusion. See Electrochemical reduction of C0 Emissions control issues, 8/ control ofC0 , 6,9 energy choice, 6 energy efficiency, 9 evaluation of Carnol process, 34, 36t, 38 greenhouse gases, 8/ renewable energy, 9 sources of C 0 , 3, 6 stationary, mobile, and natural sources of C 0 , At U.S., of C 0 from different sectors, It U.S., of C 0 from electricitygenerating units, It worldwidefromconsumption of fossil fuels, 5/ See also Catalytic oxidations Energetics carbon dioxide utilizing reactions, 55-56 C 0 transformation to liquid fuels, 167 See also Computational analysis of energy of methane reforming; Gibbsfreeenergy Energy choice, C 0 emission control, 6 Energy efficiency, C 0 emission control, 9 Energy recovery, 24 Energy requirements, barrier for C 0 conversion and utilization, 20 Energy supply, C 0 formation, 3 Environmentally benign solvents 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

2

2

2

2

2

2

2

2

expanding room temperature ionic liquids, 383 supercritical and near-critical dense carbon dioxide, 364-365 See also Dense phase carbon dioxide Epoxides. See Copolymerization of C 0 with epoxides Esterification, dense phase C 0 , 371,373 Ethane addition to C0 -reforming of methane, 135 oxidative dehydrogenation to ethylene, 45-46 oxidative dehydrogenation with C0 , 49, 50/ See also Methane conversion Ethylbenzene oxidation of derivatives, 122— 123 See also Gas-phase 0 oxidation of alkylaromatics Ethylene coupling methane, 41-42 oxidative dehydrogenation of ethane, 45-46 See also Methane conversion Ethylene oxidation selectivity, 40/ See also Catalytic oxidations Extraction. See Supercritical extraction 2

2

2

2

2

F Fe/Mo/DBH catalyst. See Gasphase 0 oxidation of alkylaromatics Fischer-Tropsch synthesis catalysts, hydrocarbons and oxygen-containing compounds, 147 Fixation 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

414 greenhouse gas concentration in atmosphere, 345 role of metal in C 0 , 56, 60 Fluidized- and fixed-bed reactors illustration of fluidized, 306/ See also Methane reforming with C 0 and 0 ; Methane reforming with C 0 over Ni/Si0 -MgO Fluid mixtures, critical phase behavior, 367 Formation, routes for C 0 , 3 Fuel production. See Carnol process Fuels, liquid, U.S. production in 1999, 17/ 2

2

2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

chemical modification of CVD Fe/Mo/DBH with Ag-doping, Ag/Fe/Mo/DBH, 121 chemical modification of CVD Fe/Mo/DBH with silica deposition, Si/Fe/Mo/DBH, 120-121 controlling molar ratio of terephthaldehyde to ptolualdehyde, 114 co-presence of C 0 in 0 feed stream, 113 effect of C 0 on /^-xylene over CVD Fe/Mo/DBH vs. reaction temperature, 116/ /?-ethylbenzene and styrene in C 0 plus 0 over CVD Fe/Mo/DBH, 122/ ethylbenzene derivatives, 122123 experimental results on p-xylene oxidation, 117/ mechanism, 123-125 peroxocarbonate intermediate, 124-125 polymethylbenzenes, 117 preparation of CVD Fe/Mo/DBH, 113 selectivity of terephthaldehyde and /?-tolualdehyde, 117, 119/ synergistic interaction of catalytic species with DBH, 123-124 p-xylene over Ag/Fe/Mo/DBH, 121/ p-xylene over CVD Fe/Mo/DBH, 113-117 /^-xylene over DBH host matrix, 124/ /?-xylene over silica modified catalyst, CVD Si/Fe/Mo/DBH, 120/ Geologic formation basis for approach as slurry reactor, 176-178 effort at Brookhaven National 2

2

2

Gas-diffusion electrodes. See Electrochemical reduction of C0 Gasoline and olefins CO- and C0 -rich syngas, 147, 149 hydrocarbon synthesis from syngas, 148/ selective synthesis from C 0 - H mixture, 147-149 See also Catalytic conversions Gas-phase 0 oxidation of alkylaromatics active catalytic sites form terephthaldehyde and ptolualdehyde formation, 114, 115 alkylaromatics over chemical vapor deposition (CVD) Fe/Mo/DBH (deboronated borosilicate molecular sieve HAMS-1B-3) catalyst, 113 anaerobic condition in presence of C 0 without 0 in feed stream, 118/ 2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

415 Laboratory (BNL) and Pennsylvania State University (PSU), 178-179 evidence for catalytic effects of certain metals in naturally occurring rocks, 177-178 oil and gas reservoirs, 176 temperature and pressure gradients of earth, 176-177 See also Catalytic reduction of C0 Gibbs free energy elimination pathways from metallo-carboxylate system, 59 formation, 13/ thermodynamics of C 0 conversion, 11, 14 Global warming, C 0 methanation, 136 Green chemistry, C 0 attracting attention, 112 Greenhouse gas (GHG) carbon dioxide, 3 emissions, 8/ methane reforming with C 0 , 197-198

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

2

2

catalysts using s c C 0 , 370371,372/ dense phase C 0 , 369-370 effect of G a 0 addition to four component catalyst, 140 first pilot test plant, 138 integrated approach to catalytic, 169-175 methanol synthesis activity, 138, 140 methanol synthesis by C 0 , 138— 140 methanol synthesis for C 0 - and CO-rich syngas on fourcomponent catalyst, 141/ methanol synthesis from C 0 on various catalysts, 139/ Pd-Ga-modified catalyst for C O rich syngas conversion, 140 See also Catalytic conversion; Methanol synthesis from C 0 containing syngas 2

2

2

3

2

2

2

2

2

H

Insertion reaction, C 0 into C - C bonds, 62, 63 Investment incentives, barrier for C 0 conversion and utilization, 20 Iridium-loaded catalysts, methane reforming o f C 0 , 207 Isobutylene, oxidative dimerization to 2,5-dimethylhexadiene (DMH), 42 Isocyanates, synthesis, 65-66 2

2

Heterogeneous catalytic reactions, dense phase C 0 , 367-373 High pressure. See Pressure effects Hydroformylation, dense phase C 0 , 371,373 Hydrogen production by Carnol process, 32 reaction with carbon dioxide, 33 Hydrogénation C u - Z n - C r - A l mixed oxide catalyst by uniform gelation method, 138 cyclohexene on supported Pd 2

2

2

Κ

Kinetics, dense phase C 0 , 366367 Kolbe-Schmitt reaction, carboxylation, 66-67 2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

416 C 0 conversion, 56-60 oxidation, 123-125 See also Synthetic chemistry Metal oxides. See Methane conversion Metals catalysis in butadiene-C0 chemistry, 62 elimination pathways from metallo-carboxylate system, 59 exchange reaction between olefin and N i - C 0 complex, 58 free energy change, 59 modes of interaction of M-olefin system with C 0 , 58 role in C 0 fixation, 56, 60 See also Synthetic chemistry Methanation reactions activity for Ni-La 0 -Ru catalyst, 137/ applying Ni-La 0 -Ru catalyst higher space velocity of reaction gas, 137-138 global warming, 136 Ni-based three-component catalyst, 136-137 rapid C0 -, 135-138 review, 136 Methane conversion amounts of C 0 chemisorbed on Ca binary catalysts after reaction, 99/ C yield over Ca-Cr and Ca-Mn catalysts, 91,93/ catalyst characterization, 88-89 catalyst composition, 89, 91 catalyst materials and preparation, 86, 88 catalytic runs and product analysis, 88 chemisorption of C 0 and catalyst state, 96, 98 C 0 hydrogénation, 138-140 coupling to ethylene, 41-^2 dependence of, and C selectivity 2

La 0 catalysts, Ru-loaded. See Methane reforming with C 0 Liquid fuels Carnol process, 33-34 liquid phase catalytic hydrogénation ofC0 /CO, 174/ methanol and higher oxygenates as automotive, 33-34 proposed integrated system for producing ultra clean, from C 0 , 169, 170/ U.S. production in 1999, 17/ See also Catalytic reduction of C0 Low temperature methane decomposition. See Methane decomposition Lummus process ammoxidation of o-xylene to ophalonitrile, 41 See also Catalytic oxidations 2

3

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

2

M

2

2

2

2

2

2

3

3

2

Maleic anhydride, butane oxidation to, 42-43 Manganese catalysts. See Methanol synthesisfromC 0 containing syngas Manufacturing industrial products, C 0 formation, 3 Market, potential, C 0 utilization, 14, 16 Market size, barrier for C 0 conversion and utilization, 20 Mars-van Krevelen oxidation of butane to maleic anhydride, 41/ See also Catalytic oxidations Mass transfer resistances, eliminating inter-phase, 369-373 Mechanism 2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

417 on partial pressure of C 0 over Ca-Ce catalyst, 91, 95/ 96 dependence of, and C selectivity on partial pressure of C 0 over Ca-Cr or Ca-Mn, 96, 97/ dependence of yields of C H and C H at equilibrium on C 0 / C H ratio and temperature, 87/ effect of catalyst composition on, with binary Ca-Cr and Ca-Mn catalysts, 91,92/ effect of catalyst composition on performance of binary Ca-Ce catalysts, 89, 90/ effect of time on stream of C yield over Ca binary catalysts, 91,95/ equations for formation of C hydrocarbons, 86 formation rates of products in reaction of C H alone with CaCe and Ca-Cr oxides, 94/ partial pressure of C0 , 91, 96 profiles for C 0 desorption during temperature programmed desorption (TPD) measurements in different atmospheres of Ca-Ce catalyst, 97/ proposed reaction mechanisms, 98, 100 relationship between C H conversion and C selectivity in reaction with various metal oxides, 87/ results of catalyst characterization, 98, 99/ results of Ce0 impregnated with nitrate solutions, 89, 92/ Methane decomposition catalyst characterization, 244250 catalyst characterization methods, 243-244 2

2

2

2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

6

4

4

2

2

4

2

2

4

2

2

catalyst preparation, 242-243 catalytic activity, 250, 252 catalytic activity method, 244 characteristics and activities of catalysts, 246/ Co dispersion, 245, 247, 250 Co state and particle size of catalysts, 245, 246/ effect of reduction conditions on relative Co dispersion of catalyst, 247, 250, 251/ equation, 276 experimental apparatus, 243 H production on carbon supported catalyst at 300 or 350°C, 253/ high surface area supports, 242 hydrogénation experiments using Co supported catalyst, 252 main phases of Co in high loading catalysts, 245, 247 molar ratio of produced H to decomposed C H vs. reaction time, 254/ physical characteristics of catalysts, 244-245 pressure drop across catalyst bed, 237 profiles of concentrations of produced H and unreacted C H for all catalysts, 251/ reaction, 208,210 supported Pt, Ru and Co metal catalysts, 242 temperature programmed reduction (TPR) profiles using all catalysts, 248/ TPR patterns of hydrogen using all catalysts, 249/ typical profile of CH , C H , and C H production vs. hydrogénation time, 254/ Methane dry reforming activity and stability of catalysts, 183-184 2

2

4

2

4

3

8

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

4

4

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

418 carbide and Ni-based catalysts, 185/ catalyst preparation, 185-186 deactivating nickel catalyst, 183 experimental, 184 high surface area tungsten and molybdenum carbides, 184 initial catalytic activity of carbide, Ni-based, and noble metal catalysts, 193 molybdenum carbide at 850°C, 186, 187/ molybdenum carbide with Ti0 , 186 molybdenum carbide with T i 0 at900°C, 188/ Ni-based catalyst at 750°C, 194/ nickel-based catalysts, 190 photoelectron spectra of tungsten carbide catalyst before and after reaction, 192/ rhodium supported on alumina, 193 1% Rh on alumina at 716°C, 194/ role of operating conditions and catalyst support, 183 SEM photographs of tungsten carbide catalyst before and after reaction, 191/ thermodynamic calculations for molybdenum and tungsten carbides, 190/ tungsten carbide at 850°C, 187/ tungsten carbide characterization before and after reaction, 186, 190 X-ray diffraction pattern of tungsten carbide catalyst before and after reaction, 189/ Methane reforming with C 0 activity and stability of Ni/u-Zr0 catalysts, 199, 203 aimed reactions producing only syngas, 133 2

2

2

2

alumina, titania, and silicasupported catalysts, 208 Boudouard reaction, 208, 210 carbon formation over metal catalysts, 206-207 catalyst preparation, 198 CO/Ni catalysts, 229 conversions of methane and C 0 and selectivity to CO and H , 201/ effect of Ar and C 0 pretreatments, 217 effect of pretreatment conditions on activity over Ru/La 0 , 212/213/ effect of pretreatment conditions on activity over Ru/Y 0 , 214/ 215/ effect of pretreatment on activity of noble metal-loaded Y 0 catalysts, 217, 221/ effect of pretreatment on activity over metal oxide supported Ru catalysts, 210, 216/ effect of time on stream (TOS) on methane conversion over Ruloaded catalysts, 208, 209/ endothermic heat of reaction, 132 equation, 276 ethane or propane addition, 135 experimental, 198, 207-208 influence of Co/Ni ratio in catalyst on conversion, H selectivity, and H /CO product ratio, 231/ influence of Co/Ni ratio in catalyst on pressure drop across catalyst bed, 230/ influence of space velocity (GHSV) on methane conversion over 27%Ni/u-Zr0 , 199, 201/ influence of temperature, space velocity and C H / C 0 feed ratio, 232/ 2

2

2

2

2

3

3

2

3

2

2

2

4

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

419 methane and C 0 conversions vs. TOS over u-Zr0 , 202/ methane conversion vs. reaction TOS over Ni catalysts, 200/ methane conversion vs. TOS over Ni/u-Zr0 catalysts with different Ni loadings, 200/ methane decomposition, 208, 210 Ni-based three-component catalyst, 132-133 performance of iridium-loaded catalysts, 207 producing synthesis gas, 197-198 product distributions for methane conversion over Ru-loaded catalysts, 208, 209/ rapid, 131-135 reactions for production of synthesis gas, 206 reviews, 132 Rh addition to Ni-based threecomponent catalyst, 133,134/ Ru-loaded Y 0 and La 0 catalysts, 210 synergistic effect of composite catalysts and combined reactions, 132-133 transition metals (group VIII) on metal oxides as catalysts, 206 XRD analyses of Ru/La 0 catalyst after reaction, 211/ XRD analyses of Ru/Y 0 and Ru/Al 0 catalysts before and after Ar or C 0 pretreatments, 218/219/220/ See also Catalytic conversion; Cobalt-nickel oxide catalysts; Computational analysis of energy of methane reforming; Simultaneous C 0 and steam reforming of methane Methane reforming with C 0 and 0 comparison of C H and C 0 2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

3

2

3

2

2

2

3

3

3

2

2

2

2

4

2

conversion, and H /CO between fluidized and fixed bed, 307/ dependence of C H and C 0 conversion, H /CO ratio and temperature difference on space velocity, 308/ 309 dependence of C H conversion on space velocity using fluidized bed and fixed bed reactor, 312-313 dependence of conversion and H /CO ratio on total pressure 2

4

2

2

4

2

over Nio.03Mgo.97O, 310

effect of introduction method on C H conversion vs. space velocity using fixed bed reactor, 311/312 effect of oxygen concentration on conversion, H /CO ratio, and temperature difference, 309310 effect of reaction temperature on C H and C 0 conversion and H /CO ratio, 307/ 308 experimental, 305 heat supply problem, 304 illustration of fluidized bed reactors, 306/ internal heat supply using fluidized bed reactor, 304 model of fluidized bed reactor, 313-314 Ni Mgi_ O catalyst preparation by coprecipitation, 305 pressurized syngas, 304 reforming method, 305 reforming under atmospheric pressure condition, 306-310 reforming under pressurized condition, 310-314 temperature gradient in catalyst bed, 304 See also Computational analysis of energy of methane reforming 4

2

4

2

2

x

x

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

420 Methane reforming with C 0 over Ni/Na-Y and N i / A l 0 activity and performance of commercial catalysts, 270, 27 k, 272 analysis of used Ni/Na-Y and N i / A l 0 after reaction, 266267 carbon monoxide disproportionation, 269-270 catalyst Ni/Na-Y and N i / A l 0 preparation, 259 C 0 conversion as function of time on stream (TOS) for catalysts, 261,262/ computational analysis of thermodynamics, 270 conversion and product yields, 263/ effect of N i loading in Ni/Na-Y catalyst on C H and C 0 conversion, 265-266 effect of reaction pressure on C 0 and C H conversion on Ni/NaY , 264/ effect of reaction temperature on C 0 and C H conversion on Ni/Na-Y, 263/ experimental, 259-260 filamentous and amorphous carbon on catalyst surface, 269 monitoring catalyst bed temperature, 260 Ni/Na-Y by incipient wetness impregnation (IWI), 266 origin of carbon formation, 269 physicochemical properties of N i catalysts supported on Na-Y and γ-Α1 0 , 260-261 preferring high pressure conditions for industrial operation, 258-259 reaction conditions, 259-260 resistance of Ni/MgO or Ni/CaO to coke formation, 270 2

2

2

3

3

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

2

3

2

4

4

2

4

2

3

S E M photographs of N i catalysts before and after use, 267, 268/ 269 surface area determinations, 259 temperature programmed oxidation (TPO) for catalysts before and after reaction, 267 Methane reforming with C 0 over Ni/Si0 -MgO apparatus and conditions for reforming, 277-278 carbonaceous species along fixed-bed, 285, 287 catalyst preparation and characterization, 277 catalytic activity using fluidizedand fixed-bed reactors, 278, 280-287 characterization of nature of coke, 282, 285 coke determination, 278 conditions for fluidized- and fixed-bed experiments, 278 conversions as function of space velocity using fixed-bed reactor, 279/ deposited coke, 282,285, 287 effects of catalyst-bed temperature, 280, 281/ 282 equations, 276 evidence of rapid catalyst deactivation in fixed-bed, 282, 283/ fluidized-bed reactor use, 276 limitations of fixed-bed reactor, 276 profiles of pressure-drop, 282 reforming results, 280/, 281/ X P S analysis of fresh and spent catalyst, 284/ X P S relative intensities of crystallized to amorphous forms and surface elemental compositions of catalyst, 285/ XPS spectra of top and bottom

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

421 recovered catalyst samples, 286/ Methane reforming with C 0 over Rh/Na-Y and R h / A l 0 carbon types on coked catalysts, 300 catalyst preparation by incipient wetness impregnation (IWI), 290 C 0 conversion for atmosphericand high-pressure, 293/ C O and H yields at atmosphericand high-pressure, 294/ coking equations, 300 conversion and product yields, 293/ conversion and product yields for atmospheric- and high-pressure, over Rh/Na-Y (ion exchange, IE) and Rh/Na-Y (IWI), 296297 decrease in activity between atmospheric- and high-pressure operations, 292, 294 effect of Rh loading in catalysts by IWI, 297-298 experimental, 290-291 H and C O pulse chemisorption of Rh catalysts before and after, 294-295 metals favoring coking, 300-301 performance of supported N i catalysts for atmospheric- and high-pressure, 295 physico-chemical properties of catalysts, 291,292/ reaction conditions, 290-291 resistance of noble metals to coking, 290 Rh catalysts and deactivation, 294 Rh/Na-Y by IE preparation vs. Rh catalysts on amorphous supports, 296 2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

3

S E M photographs of fresh and used catalysts, 298-299 surface area determination, 290 temperature programmed oxidation (TPO) of used catalysts, 299-300 Methanol catalyst design considerations, 173, 175 liquid automotive fuel, 33-34 Methanol synthesis from C 0 containing syngas activity of C u - M n oxide on commercial support, 159/ catalysis of bulk C u - M n oxide, 156-158 catalysis of C u - M n - 0 / T i 0 , 162 catalysis of C u - M n - 0 / Z r 0 , 160 catalysis of supported C u - M n oxide, 158, 160, 162 comparison of C u - M n - 0 loading effect, 164/ C u - M n synergy on T i 0 , 162/ C u - M n synergy on Z r 0 , 161/ effect of C 0 on activity of C u M n oxide bulk catalyst, 157158 effect of C u - M n - 0 loading on T i 0 , 164/ effect of C u - M n - 0 loading on Z r 0 , 160/ effect of Cu/Mn ratio on activity and surface area, 156 experimental, 154-156 list of commercial support, 154/ preparation of high surface area support, 155 preparation of T i 0 support, 159/ preparation of Z r 0 and T i 0 with high surface area, 158, 160 STY and surface area as function of Cu content of C u - M n oxide, 156/ 2

2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

422

STY from CO/H as function of Cu content of Cu-Mn oxide, 157/ synergistic effect of Cu and Mn, 153-154 XRD of Cu-Mn-0 on Ti0 , 163/ XRD of Cu-Mn-0 on Zr0 , 161/ See also Catalytic conversions; Hydrogénation; Olefins and gasoline Methyl methacrylate, new process, 45 Methyl nitrite, dimethyl carbonate (DMC) synthesis, 72 Mitigation technology. See Carnol process; Catalytic conversions Molybdenum carbide dry reforming of methane at 850°C,187/ preparation and methane dry reforming, 186 thermodynamic calculations, 190/ with Ti0 , dry reforming of methane at 900°C, 188/ See also Methane dry reforming Multi-functional catalysts. See Catalytic conversions 2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

with C 0 ; Methane reforming with C 0 and 0 Ni/Na-Y and Ni/Al 0 catalysts. See Methane reforming with C 0 over Ni/Na-Y and Ni/Al 0 Nylon intermediate manufacturing, dinitrogen oxide, 47 2

2

2

2

3

2

Ο Olefin epoxidation, oxidation in dense phase C 0 , 374, 376 Olefins, epoxidation using Jacobson's catalyst, 376 Olefins and gasoline CO- and C0 -rich syngas, 147, 149 hydrocarbon synthesis from syngas, 148/ selective synthesis from C 0 - H mixture, 147-149 See also Catalytic conversions Oxidants cost of oxygen sources, 46, 47/ N 0 as selective for aromatics, 46-47 N 0 in nylon intermediates manufacturing, 47 proposed mechanism for N 0 oxidation of benzene, 47/ titanosilicates, 47-^8 Oxidations dense phase C 0 , 371, 373 homogeneous, in scC0 , 373-377 See also Catalytic oxidations; Gas-phase 0 oxidation of alkylaromatics Oxygen miscibility with scC0 , 376 See also Methane reforming with C 0 and 0 Oxygenates, methanol and higher, liquid automotive fuel, 33-34 2

2

2

2

2

2

Ν Nickel catalysts addition of cobalt for conversion of methane to syngas, 225 commercial catalyst testing, 190

2

2

2

deactivation, 183 dry reforming of methane at 750°C, 194/ methane dry reforming, 185 See also Cobalt-nickel oxide catalysts; Methane dry reforming; Methane reforming

2

3

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

423 irradiation in presence of C 0 and H 0,333/ photoluminescence spectrum of Ti-oxide anchored on zeolite, 336/ physical properties and photocatalytic activity of T i 0 catalysts, 333/ product distribution of large-pore Ti-containing zeolite, Ti-Beta, 339 product distributions of powdered T i 0 and Ti-oxide/Y-zeolite catalysts, 336-337 reaction scheme, 331/ small particle T i 0 catalysts, 332-333 Ti-containing zeolite and mesoporous molecular sieves, 337-339 T i 0 single crystals, 334-335 Ti-oxide anchored in zeolite (ion exchange), 335-337 Ti-oxide anchored on porous silica glass (CVD), 340 Ti/Si binary oxide (sol-gel), 341 titanium oxide catalysts, 331-332 utilization of solar energy, 330331 XANES and Fourier transformation of EXAFS spectra of Ti-oxide/Y-zeolites, 335/ yields of CH OH and C H per unit weight Ti-based catalysts, 338/ yields of formation of C H and CH OH in, with water, 334/ See also Catalytic reduction of C0 Physical adsorption, separation of gas mixtures, 10 Platinum catalysts, Pt-loaded T i 0 powder, 333-334 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

Paraffin analytical of extraction of asphalts, 394-395 asphalt constituents, 391-392 extractionfromasphalt, 395-398 extraction yieldsfromasphalt, 399 incentive for direct oxidation, 43/ oxidations, 43-44 production, 390 See also Asphalt Pennsylvania State University (PSU), integrating C 0 sequestration with conversion in geologic formations, 178-179 Petroleum industry applications of supercritical extraction, 388 C 0 production, 388 deasphalting, 389-390 tailoring analytical methods, 392 See also Asphalt; Deasphalting process; Paraffin Phenol, enzymatic conversion to 4OH-benzoate, 67 Phosgene, dimethyl carbonate (DMC) synthesis, 72 Photocatalytic reduction of C 0 Cu-loaded and Pt-loaded T i 0 powder, 333-334 design of systems, 330 effect of Pt-loading on photocatalytic activity of Ticontaining zeolite, 338-339 effects of number of Ti-0 layers of anchored Ti-oxide catalysts, 340/ effects of T i 0 contents on yields of C H and relative intensity of photoluminescence of Ti/Si binary oxide catalysts, 341/ ESR signals with T i 0 under UV2

2

2

2

2

4

2

2

2

2

2

3

4

4

3

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

2

424 Platinum/gas-diffusion electrode. See Electrochemical reduction of

co

2

Polymethylbenzenes. See Gasphase 0 oxidation of alkylaromatics Pore-diffusion resistances, alleviation of internal, 368-369 Pressure effects dense phase C 0 , 366-367 See also Computational analysis of energy of methane reforming; Methane reforming with C 0 and 0 ; Methane reforming with C 0 over Ni/Na-Y and Ni/Al 0 ; Methane reforming with C 0 over Rh/Na-Y and Rh/Al 0 Pretreatments. See Methane reforming with C 0 Propane addition to C0 -reforming of methane, 135 ammoxidation to acrylonitrile, 44/ oxidation to acrylic acid, 44/ Propylene, epoxidation to propylene oxide, 48, 49/ Propylene ammoxidation selectivity, 40/ See also Catalytic oxidations Propylene oxidation acrolein and acrylic acid, 42 selectivity, 40/ See also Catalytic oxidations Propylene oxide. See Copolymerization of C 0 with epoxides 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

2

2

2

2

3

2

2

3

2

2

2

Recycling C0 , 24 sequestration option, 167 Reduction. See Catalytic reduction of C0 ; Electrochemical reduction of C 0 ; Photocatalytic reduction of C 0 Renewable energy, C 0 emission control, 9 Research, utilization of C 0 , 2124 Residue oil supercritical extraction (ROSE), industrial process, 389 Reverse water-gas-shift (RWGS) reaction, equation, 276 Rh catalysts. See Methane reforming with C 0 over Rh/NaY and Rh/Al 0 Rhodium on alumina dry reforming of methane at 716°C, 194/ noble metal catalyst for dry reforming of methane, 193 See also Methane dry reforming Ruthenium-loaded catalysts. See Methane reforming with C 0 2

2

2

2

2

2

2

2

3

2

S Sequestration C 0 , 9-10 effort at Brookhaven National Laboratory (BNL) and Pennsylvania State University (PSU), 178-179 scheme for aquifer/ocean, of C0 , 169, 170/ subterranean C 0 options, 167 See also Catalytic reduction of C0 Silica glass, Ti-oxide anchored on porous, 340 Simultaneous C 0 and steam reforming of methane 2

2

2

R

2

Reaction heat. See Computational analysis of energy of methane reforming

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

425 catalyst performance at optimum Co/Ni ratio, 235/ 236/ Co/Ni catalysts, 233 influence of Co/Ni ratio on catalyst performance, 234/ See also Cobalt-nickel oxide catalysts Solar energy, utilization, 330-331 Solubility C 0 studies, 169, 171-173 See also Catalytic reduction of 2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

co

2

Supported copper and manganese. See Methanol synthesis from C 0 containing syngas Syngas methane reforming with C 0 , 197-198 methanol synthesesfromC 0 and CO-rich, on fourcomponent catalyst, 140, 141/ pressurized, 304 selective synthesis of light olefins and gasoline from C 0 - H mix, 147-149 See also Catalytic conversions; Methanol synthesisfromC 0 containing syngas Synthesis. See Dimethyl carbonate (DMC) Synthetic chemistry biotechnological synthesis of 4OH-benzoic acid, 68 carboxylation of organic substrates, 60-62 carboxylation reactions, 58 coupling chemistry with biotechnology, 66-68 coupling of two C 0 molecules, 61 development of carbon dioxidebased industry, 55 elimination pathways form metallo-carboxylate system, 59 energetics of carbon dioxide utilizing reactions, 55-56 enzymatic conversion of phenol to 4-OH-benzoate, 67 exchange reaction occurring when olefin reacts with N i - C 0 complex, 58 free energy change of elimination step, 59 insertion of C 0 into C - C bonds, 62, 63 Kolbe-Schmitt reaction, 67 2

2

2

2

Solvent expansion addition ofC0 , 378 ionic liquids with scC0 , 383 Solvents, environmentally benign chemical processing, 364 Steam reforming of methane. See Simultaneous C 0 and steam reforming of methane Stilbene, oxidative dimerization of toluene, 42 Styrene. See Gas-phase 0 oxidation of alkylaromatics Supercritical alkylation, dense phase C0 , 368-369 Supercritical C 0 limitations, 377-378 solvent expansion, 378, 383 See also Dense phase carbon dioxide Supercritical extraction analytical of paraffin extraction of asphalts, 394-395 C 0 , 23 experimental results, 395-398 extraction conditions, 393-394 selectivity, 390-391 solvating capacity of C0 , 388— 389 See also Asphalt; Paraffin; Petroleum industry Supercriticalfluids,applications in processing materials, 389 2

2

2

2

2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

426 mechanism of C 0 conversion, 56-60 metal catalysis in butadiene-C0 chemistry, 61-62 modes of interaction of M-olefin with C 0 , 58 reactivity of coordinated C 0 towards electrophile, 57 reactivity of cycloalkanes towards C 0 , 63 reduction reactions, 56 role of metal in C 0 fixation, 56, 60 synthesis of carbamates and isocyanates, 65-66 synthesis of carbonates, 62, 6465 transesterification reactions, 66 utilization of carbon dioxide, 55 Synthetic plastics, U.S. production in 1999, 15/ 2

2

2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

Temperature effects. See Computational analysis of energy of methane reforming Temperature programmed oxidation (TPO), used Rh catalysts, 299-300 Temperature programmed reduction (TPR) curves of catalyst with different Co/Ni ratios, 227, 228/ dependence of trend of curves on Co/Ni catalyst ratio, 233, 237 methane decomposition on supported Co catalysts, 247, 248/ patterns of hydrogen production for supported Co catalysts, 247, 249/ surface area and degree of

reduction of Co/Ni supported catalyst, 227/ Thermodynamics calculations for molybdenum and tungsten carbides, 190/ Carnol process, 35/ C 0 conversion, 11,14 C 0 transformation to liquid fuels, 167 dense phase C 0 , 366-367 Gibbsfreeenergy of formation, 13/ See also Computational analysis of energy of methane reforming Ti/Si binary oxide, 341 Titania catalysts catalysis of Cu-Mn-0/Ti0 for methanol synthesis, 162 comparing Cu-Mn-0 loading effect on zirconia and, 164/ Cu-Mn synergy, 162/ effect of Cu-Mn-0 loading, 164/ preparation with high surface area, 158, 160 XRD of Cu-Mn-0 on, 163/ See also Methanol synthesis from C 0 containing syngas Titanium oxide catalysts Cu-loaded and Pt-loaded T i 0 powder, 333-334 small particle T i 0 catalysts, 332-333 Ti-containing zeolite and mesoporous molecular sieves, 337-339 T i 0 single crystals, 334-335 Ti-oxide anchored on porous silica glass, 340 Ti-oxide anchored on zeolite, 335-337 Ti/Si binary oxide (sol-gel), 341 See also Photocatalytic reduction ofC0 Titanosilicates 2

2

2

2

2

2

2

2

2

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

427 epoxidation of propylene to propylene oxide, 48, 49/ oxidations, 47^8 Toluene, oxidative dimerization to stilbene, 42 Transesterification, isocyanate synthesis, 65-66 Tungsten carbide dry reforming of methane at 850°C,187/ photoelectron spectra before and after reaction, 192/ preparation and methane dry reforming, 185-186 SEM photographs of, before and after reaction, 191/ thermodynamic calculations, 190/ X-ray diffraction pattern before and after reaction, 189/ See also Methane dry reforming

replacing hazardous or less effective substances, 23-24 solar energy, 330-331 strategies for research on C 0 , 21-24 supercritical extraction, 23 with and without chemical conversion processing, 22/ 2

See also Copolymerization of

C 0 with epoxides 2

W Water. See Photocatalytic reduction ofC0 2

X o-Xylene, ammoxidation to ophalonitrile, 41 Xylenes. See Gas-phase 0 oxidation of alkylaromatics 2

U Utilization annual U.S. production of C 0 and related chemicals (1999), 15/ barriers for C0 , 20-21 challenges and strategies for C 0 , 16, 20-24 chemical processes, 19/ C 0 , 10-11 C 0 as co-reactant or co-feed, 21 converting C 0 under geologicformation conditions, 24 copolymerization of C 0 with epoxides, 103 current status and potential market of C 0 , 14, 16 energy recovery, 24 lack of socio-economical and political driving forces, 21 recycling C 0 , 24 2

2

2

2

2

Y Y 0 catalysts, Ru-loaded. See Methane reforming with C 0 Yttrium-metal coordination catalyst copolymerization results for different catalyst systems, 105, 106/ preparation, 104 2

3

2

See also Copolymerization of

2

C 0 with epoxides 2

2

Ζ

2

2

Zeolite Ti-containing and mesoporous molecular sieves, 337-339 Ti-oxide anchored on, 335-337

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.

428 See also Titanium oxide catalysts Zirconia catalysts catalysis of Cu-Mn-0/Zr0 for methanol synthesis, 160 comparing Cu-Mn-O loading effect on titania and, 164/ Cu-Mn synergy, 161/ dimethyl carbonate (DMC) formation, 73-74

3

4

2

2

2

Downloaded by 217.66.152.143 on September 14, 2012 | http://pubs.acs.org Publication Date: January 24, 2002 | doi: 10.1021/bk-2002-0809.ix002

2

effect of H P0 /Zr0 preparation method on DMC formation, 81, 83/ preparation, 74 preparation with high surface area, 158, 160 See also Dimethyl carbonate (DMC); Methane reforming with C 0 ; Methanol synthesis from C 0 containing syngas

In CO2 Conversion and Utilization; Song, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2002.